WO2014203170A1

WO2014203170A1 - Estimating utilization of a space over time

Info

Publication number: WO2014203170A1
Application number: PCT/IB2014/062307
Authority: WO
Inventors: Michael GRESTY; Irina Mladenova
Original assignee: Gresty Michael; Irina Mladenova
Priority date: 2013-06-17
Filing date: 2014-06-17
Publication date: 2014-12-24

Abstract

Described herein are utilization analysis facilities that analyze information indicative of a multitude of electronic communication signals detected in a space over time and, based on that information, determine inferences about utilization of that space during a time period of interest. In some embodiments, for example, the facilities analyze electronic communication signals detected in a larger space over a large period of time and, using the analysis techniques described herein on those signals, identify utilization of a portion of that space, or multiple different portions of that space, during a time period of interest. Also described herein are examples of utilization analysis facilities that analyze information indicating requests for access for a space and, based on that information, predict times at which occupants will determine a space. Based on the requests for access and the predicted departures, the facilities identify utilization of the space.

Description

ESTIMATING UTILIZATION OF A SPACE OVER TIME

CROSS REFERENCE TO RELATED APPLICATIONS

The present application claims priority under 35 U.S.C. § 119(e) to U.S. provisional patent application serial number 61/835,760, filed June 17, 2013, the entire contents of which are hereby incorporated by reference herein. To the extent that any terminology used in the provisional patent application is inconsistent with the usage of that terminology herein, the terminology should be understood in a manner most consistent with discussion below. BACKGROUND

Offices and office buildings are typically designed with a certain capacity in mind. For example, a business may design an office to house a certain number of employees. To do so, the business may design the office to house a desk for each employee, such as a certain number of cubicles and personal offices. The business may also design the office to include conference rooms of particular sizes, based on what the business believes will be its usual needs for meeting spaces.

Such businesses may be interested in determining whether their offices are appropriately accommodating their needs, such as whether the office is too small or too large for the business. A business may do this by comparing the number of current employees of the business to the number of current desks in the office. This technique may tell the business whether its office can house all of its employees at once. However, all of the business' employees may seldom be in the office at once, due to employee absences, employees working remotely, etc. Another technique a business may use to determine whether its office is accommodating appropriately its needs on a particular day is to review a number of employees that came to work that day, which it may determine using "swipes" of employee ID badges at the front entrance of the office. A "swipe" is a presentation of an employee credential by an employee to a device that scans the credential to determine whether the employee is authorized to do something. Bar code scanners or RFID scanners could be used to scan the credentials, and the credentials could be employee ID badges. By comparing the individual employees identified from those swipes to the number of desks in the office, the business tries to determine how well the office met the business' needs that day and (assuming the day is a typical day) how well the office typically meets the business' needs. SUMMARY

In one embodiment, there is provided a method for tracking utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space. The method comprises operating at least one programmed processor to perform an act comprising receiving information indicating at least a first set of electronic communication signals detected in the space over a first time period. The first set of electronic communication signals having been emitted by a first set of devices, where the information indicates at least, for each detected electronic communication signal, an identifier for one of the first set of devices that emitted that detected electronic communication signal, an estimated location of the one device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected. The method further comprises operating the at least one programmed processor to determine, based on the information indicating the set of electronic communication signals emitted by the first set of devices, a limited set of devices that are each uniquely associated, among the first set of devices, with an occupant of the space. The limited set of devices is a subset of the first set of devices. The determining the limited set that are uniquely associated comprises, in a case that two or more devices of the first set of devices are determined to be associated with one occupant, identifying only one of the two or more devices to include in the limited set of devices. The method further comprises operating the at least one programmed processor to determine utilization of the space during the time period of interest based at least in part on information indicating a second set of electronic communication signals that were detected as having been emitted by devices of the limited set of devices.

In another embodiment, there is provided at least one computer-readable storage medium having encoded thereon executable instructions that, when executed by at least one computer, cause the at least one computer to carry out a method for enabling tracking of utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space. The method comprises receiving information indicating electronic communication signals detected in the space over a first time period different from the time period of interest. The electronic communication signals detected in the space were emitted by a first set of devices. The information indicates at least, for each detected electronic

communication signal, an identifier for a device that emitted that detected electronic communication signal, an estimated location in the space of a device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected. The method further comprises determining, based at least in part on the information indicating the electronic communication signals detected in the space over the first time period, that at least a first device and a second device of the first set of devices that emitted electronic communication signals indicated in the information are associated with one occupant of the space, and evaluating utilization of the space during the time period of interest based on electronic communication signals detected in the space at least in part by filtering, from second information indicating electronic communication signals detected in the space during the time period of interest, electronic communication signals emitted by the second device.

In a further embodiment, there is provided an apparatus comprising at least one processor and at least one storage medium having encoded thereon executable instructions that, when executed by the at least one processor, cause the at least one processor to carry out a method for enabling tracking of utilization of a space by occupants through evaluating electronic communication signals detected in the space. The method comprises receiving first information indicating electronic communication signals detected in the space over a time period. The electronic communication signals detected in the space were emitted by a first set of devices. The information indicates at least, for each detected electronic communication signal, an identifier for a device that emitted that detected electronic communication signal, an estimated location in the space of a device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected. The method further comprises determining, based at least in part on the information indicating the electronic communication signals emitted by the first set of devices and detected in the space over the first time period, a first set of occupants that operate the first set of devices. A number of occupants in the first set of occupants is less than a number of devices in the first set of devices. The method further comprises, based at least in part on the estimated locations of the electronic communication signals indicated by the information, identifying movements of each occupant of the first set of occupants during the time period and, based on the movements, classifying each of the occupants in the first set of occupants into one of a set of mobility classifications. The mobility classifications of the set indicate an effect of an occupant on occupancy of the space.

In another embodiment, there is provided at least one computer-readable storage medium having encoded thereon executable instructions that, when executed by at least one processor, cause the at least one processor to carry out a method for enabling tracking of utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space. The method comprises receiving information indicating electronic communication signals detected in the space over a first time period different from the time period of interest. The electronic communication signals detected in the space were emitted by a first set of devices. The information indicates at least, for each detected electronic communication signal, an identifier for a device that emitted that detected electronic

communication signal, an estimated location in the space of a device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected. The method further comprises identifying at least one device of the first set of devices for which emission of communication signals by the at least one device during the time period of interest is not to be considered in evaluating utilization of the space during the time period of interest. Identifying the at least one device comprises evaluating the information indicating the electronic communication signals detected in the space over the first time period. The method further comprises evaluating utilization of the space during the time period of interest at least in part by filtering, from second information indicating electronic communication signals detected in the space during the time period of interest, electronic communication signals emitted by the at least one device.

In a further embodiment, there is provided a method for tracking utilization of a space by occupants during a time period of interest through evaluating requests for access for spaces. The requests were made by occupants over time. The method comprises operating at least one programmed processor to perform an act comprising receiving information indicating a set of requests for access for a space made by a set of occupants during a time period of interest. The information indicates for each request for access of the set at least a time that the request was made. The method further comprises operating the at least one programmed processor to predict, based on the received information indicating the set of requests for access, departure times for each occupant of the set of occupants. The departure times indicate times at which each occupant will leave the space during the time period of interest. The method further comprises operating the at least one programmed processor to determine utilization of the space during the time period of interest based at least in part on the requests for access and the departure times.

In another embodiment, there is provided a method for tracking utilization of a space by occupants during a time period of interest by evaluating electronic communication signals detected in the space. The method comprises operating at least one programmed processor to perform acts comprising receiving information indicating at least a first set of electronic communication signals detected in the space, and determining utilization of the space during the time period of interest based at least in part on the information indicating the first set of electronic communication signals detected in the space.

The foregoing is a non-limiting summary of the invention, which is defined by the attached claims. BRIEF DESCRIPTION OF DRAWINGS

The accompanying drawings are not intended to be drawn to scale. In the drawings, each identical or nearly identical component that is illustrated in various figures is represented by a like numeral. For purposes of clarity, not every component may be labeled in every drawing. In the drawings:

FIG. 1A is a flowchart of a previously-used technique for determining utilization of a space;

FIG. IB is a graph illustrating a result of the previously-used technique illustrated in FIG.

1A;

FIG. 2 is a sketch of a floor plan of an exemplary space (including sub-spaces) with which embodiments may operate, as well as examples of devices that may be included in such a space;

FIG. 3 is a flowchart of a process that may be used in some embodiments to determine utilization of a space over time;

FIG. 4 is a flowchart of a process that may be used in some embodiments to identify devices to include in a limited set to be monitored as part of identifying utilization of a space over time;

FIG. 5 is a flowchart of a process that may be used in some embodiments to identify devices that are associated with a single occupant;

FIG. 6 is a flowchart of a process that may be used in some embodiments to identify, when multiple devices are associated with one occupant/person, a primary device to be used in tracking movements of that occupant/person;

FIG. 7 is a flowchart of a process that may be used in some embodiments to determine utilization of a space during a time period of interest based on signals detected in the space over time;

FIG. 8 is a flowchart of a process that may be used in some embodiments to determine a position of a device during a time period of interest, as part of determining utilization of a space during that time period;

FIG. 9 is a flowchart of a process that may be used in some embodiments to determine utilization in accordance with input received from a user regarding types of spaces to be evaluated; FIG. 10 is a flowchart of a process that may be used in some embodiments to prepare information regarding signals detected in a space for analysis using other techniques described herein;

FIG. 11 is a flowchart of a process that may be used in some embodiments to classify the types of movements exhibited by an occupant based on signals emitted by devices used by that occupant;

FIG. 12 is a flowchart of a process that may be used in some embodiments to determine utilization of a space based on requests to access the space;

FIG. 13 is a flowchart of a process that may be used in some embodiments to determine utilization using a categorization of people based on requests to access a space made by those people;

FIG. 14 is a flowchart of a process that may be used in some embodiments to use both signals detected in a space and requests to access the space to determine utilization in the space;

FIG. 15 is a flowchart of a process that may be used in some embodiments to train, using information regarding detected signals, a process for determining utilization of a space based on requests to access the space; and

FIG. 16 is a block diagram that illustrates some components of a computing device with which some embodiments may operate.

DETAILED DESCRIPTION

A space such as an office or building typically has a chaotic electromagnetic spectrum, with many different devices wirelessly communicating within the space at one time. Those devices may include network infrastructure devices or office equipment, in addition to devices personally used by people in the space (e.g., employees that work in the office, and visitors) - which may be multiple devices per person. Further, signals detected in and near the space may have been emitted by devices that are located outside of the space, such as devices operated by people walking or driving by the office or devices in adjacent spaces such as on adjacent floors of an office building, due to communication signals from those devices penetrating into the space. Moreover, devices in a space do not communicate regularly. As a result, one device may be inside a space for a lengthy period of time without ever transmitting a signal, whereas another device in that space may transmit signals for an entirety of the time it is within the space and a third device may only intermittently transmit signals in the space over that time.

Analysis of usage of a space has previously focused on the badge swipes and other factors from which usage can be inferred, and the inventors are unaware of any prior efforts to determine utilization from an analysis of electronic communication signals, perhaps due at least in part to this low correlation between the electronic communication signals detected in a space and people using that space.

However, the inventors have recognized and appreciated that such electronic communication signals can be suggestive of utilization of a space, once complex computerized learning and statistical analysis techniques are performed on the signals. Further, the inventors have recognized and appreciated that using a specific set of complex computerized learning and analysis techniques to filter and analyze a chaotic electromagnetic spectrum may enable a finer- grain identification of occupants in a space to be made as compared to prior determination techniques. The inventors have further recognized and appreciated that having more detailed information regarding occupants of a space, and with comparison to information regarding a capacity of that space, utilization of that space over time can be determined. Such utilization may, for example, identify a number of occupants inferred to be present over a period of time as compared to a capacity of the space.

The inventors have recognized and appreciated that performing some of the computer- implemented analyses described herein on information regarding electronic communication signals emitted by devices over time may aid in the identification of a limited set of those devices. The devices in this limited set may be those that, using analysis techniques described herein, may be most highly indicative of presence of occupants in a space and of utilization of a space. Once the limited set of devices is identified using those computer-implemented analyses, specific analyses of information about usage of the devices of the limited set may then be performed to enable the fine-grained inferences regarding utilization of a space over time to be made.

Accordingly, described herein are examples of utilization analysis facilities that, when executed by one or more computing devices, may analyze information indicative of a multitude of electronic communication signals detected in a space over time and, based on that information, determine inferences about utilization of that space during a time period of interest. In some embodiments, for example, utilization analysis facilities may analyze electronic communication signals detected in a larger space over a large period of time and, using the analysis techniques described herein on those signals, identify utilization of a portion of that space, or multiple different portions of that space, during a time period of interest.

In some such embodiments, the utilization analysis facilities may collect information regarding electronic communication signals detected in a space, which may include computer- estimated locations of devices that emitted the detected electronic communication signals. From the computer-estimated locations, these utilization analysis facilities may identify corresponding locations of the devices in a space, such as by mapping the computer-estimated locations to locations within a floor plan of a space, such as to rooms within an office.

The utilization analysis facilities may also, in these embodiments, infer a relationship between devices that emitted signals detected in the space, so as to attempt to identify two or more devices that are operated by one individual in the space. The facilities may make this determination through the analysis of electronic communication signals detected in the space, and thereby infer the existence of an "occupant" that operates two or more of the devices that emitted the signals. The utilization analysis facilities may then identify a limited set of devices that are personally operated by such occupants (as opposed to being, for example, infrastructure devices) and are uniquely associated with those occupants among the set of devices that emitted the detected signals. The identification of the limited set may be made in various ways including, in some embodiments, through applying customized machine learning and statistical analysis techniques to information regarding signals detected in the space. Following identification of the limited set of devices, in these embodiments the utilization analysis facilities may review signals emitted by devices of the limited set and thereby analyze movements of the occupants. From information regarding the movements of the occupants, the utilization analysis facilities may identify spaces occupied by those occupants over time, and utilization of the space(s) over time.

While utilization analysis facilities that analyze electronic communication signals are described herein, in some embodiments utilization analysis facilities may additionally or alternatively analyze other types of data that can be indicative of presence of occupants in a space. Such other types of data may be analyzed by utilization analysis facilities to assist or supplement review of electronic communication signals or, in some embodiments, may be used for spaces for which collection of information on electronic communication signals is not possible or not feasible. Various embodiments that operate with information on data types other than data indicative of electronic communication signals to produce information on utilization of a space over time are described below.

Any suitable information that is indicative of presence of occupants in a space may be used in embodiments, as embodiments are not limited in this respect. For example, usage of network-connected devices in a space or an output of a physical occupancy sensor may indicate presence of an occupant in a space. Other embodiments that operate on data that is not indicative of electronic communication signals may analyze information on requests for access submitted by occupants for spaces. In some such embodiments, requests for access for spaces may be requests for physical access to a space. Such a request for physical access may be, for example, a request for admission to a space. A request for admission to a space may precede an occupant entering a space, and may therefore be indicative of presence of an occupant in a space and/or indicative of arrival of an occupant in the space. In other embodiments, the requests for access for a space may additionally or alternatively be requests for electronic access, such as requests to access a computer network that were submitted from within a space. In embodiments that evaluate requests for electronic access, a utilization analysis facility may analyze requests for electronic access to identify those that are submitted from within a space, which may be indicative of the presence of an occupant in that space, including the arrival of an occupant to a space.

In some embodiments that process requests for access, in addition to requests for access, occupants may also explicitly signal an end to access, which may be a signal of an occupant's departure from a physical space or a signal that the occupant is ending network access (e.g., logging off a network). In such a case, utilization analysis facilities described herein may analyze both requests for access and explicit signals of an end to access to identify utilization of a space over time. In other embodiments, however, occupants may not explicitly signal an end to access to a space. In such cases, while "arrival" of an occupant may be determined from requests for access, "departures" of the occupants are not signaled and may not be identified using traditional techniques, which complicates determining utilization of the space over time. The inventors have recognized and appreciated, however, that by analyzing the requests for access using complex computer-implemented learning and analysis techniques, inferences can be drawn regarding departures of people from a space. As such, by analyzing both the information about explicit requests for access to a space and the inferences regarding departures from a space, utilization of the space may be tracked over time.

Accordingly, also described herein are various examples of utilization analysis facilities that review information regarding presence of occupants in a space, such as requests for physical access to a space or requests for electronic access made from within a space, and may identify from that information arrivals of occupants to a space. The facilities may also make inferences regarding departures of occupants from the space, and analyze the inferred arrivals and inferred departures to track utilization of the space over time. Inferences for departures may be drawn in any of various ways, examples of which are described below. In some embodiments, for example, a utilization analysis facility may identify multiple categories of people and determine inferences about the movements of those people based on requests for physical access to one or various spaces made by those people. Subsequently, occupants of a space may be classified into one of those categories. In some such embodiments, a utilization analysis facility may then make inferences regarding departures from a space over time based on requests for access to the space and the categories of the people making those requests for access. The inferences about movements of each category of person may be made in any suitable manner, including using statistical analysis and/or machine learning techniques or, in some embodiments, based on learning about movements of individuals in general from analysis of electronic communication signals in other spaces.

In some embodiments, information indicative of the presence of an occupant in a space, such as requests for access to a space, may be used together with information on electronic communication signals detected in a space to determine utilization. For example, in some embodiments, a number of occupants in a space and utilization of the space in a time period may be determined based on electronic communication signals detected in the space in that time period. Separately, a second count of occupants in the space during the time period may be determined, such as from information indicative of presence of occupants in a space like requests for access for the space, operation of network-connected devices in a space, or outputs of physical occupancy sensors. The second count of occupants of the space may then be used to adjust the utilization determined based on the electronic communication signals. By using both sets of information, in some cases these embodiments may be able to determine a more accurate utilization of a space. While in some embodiments two such sets of information may be used, in other embodiments more than two may be used.

As discussed below, embodiments described herein are not limited to operating with a space of any particular size and may operate with any space, including in environments in which a space is composed of multiple smaller spaces (what might be termed sub-spaces), such as in the case of an office building that includes multiple floors, or a floor of an office that includes multiple personal offices. As also discussed below, embodiments described herein are not limited to operating with any specific period of time and may track utilization of a space for a day or across fractions of a day, such as by the hour or by fractions of an hour. The utilization analysis facilities may also, in some embodiments, produce reports on utilization that may be output to a user in any suitable format.

As discussed above, many of the embodiments described herein operate using information regarding electronic communication signals detected in a space. It should be appreciated that embodiments described herein are not limited to working with any particular types of device that may emit electronic communication signals, or any specific types of electronic communication signals.

In embodiments, the electronic communication signals may be electromagnetic signals. The signals may be signals transmitted within a wireless computer communication network. Additionally or alternatively, in embodiments the electronic communication signals may be signals having a transmission range greater than 10 meters and may be, for example, signals transmitted over a wireless local area network (WLAN), wireless wide area network (WW AN), wireless metropolitan area network (WMAN), or other wireless network, such as IEEE 802.11 ("WiFi") traffic. The electronic communication signals analyzed in some embodiments may include signals that do not include location-identifying data or that have data (e.g., a data payload) that does not identify a location, such as locations of devices that emitted the signals. Such signals may therefore not be location beacons. Such signals may be, for example, signals transmitting data from a source device to a destination device, wherein the destination device is distinct from a utilization analysis system and does not perform functionality to analyze electronic communication signals and/or determine utilization of a space. The destination device may be identified in a header of the signals. Such devices that serve as the source and destination of signals may be network hosts, rather than network nodes. The data signals may also have been sent in response to user operation. Accordingly, in some embodiments wireless networks may provide for the exchange of both management and control data signals that relate to operation/maintenance of the network (e.g., beacon signals) and data signals that relate to exchanging data between hosts in a network or a host in one network and a host in another network. In some such embodiments, a utilization analysis facility may process the data signals rather than the management and control signals.

The devices that emit such signals may be any suitable devices. The devices may include infrastructure devices, including networking infrastructure devices. Infrastructure devices may be devices that, though they may be operated by a user from time-to-time, are designed to be configured and subsequently operate unattended by human users. Networking infrastructure devices may be devices dedicated to performing networking-specific functionality to enable formation and maintenance of a computer network, such as wireless access points, wireless repeaters, or other networking devices. The devices may also include stationary devices.

Stationary devices may be devices that do not move during use. Infrastructure devices may also be stationary devices. The devices that emit signals processed using techniques herein may also include devices that are pieces of multi-user equipment. Multi-user equipment may be devices that are not associated with or operated by a particular user, but instead are operated by multiple users. Many types of office equipment, such as wireless printers, are multi-user equipment. Multi-user equipment may also, in many cases, by stationary devices. Devices may also include personal devices. Personal devices may be devices that are associated with and operated by a specific user. In some cases, the personal devices may be personal mobile devices, which may be devices that are arranged to be carried by the associated user during normal user. Personal devices may include personal computing devices, and personal mobile devices may include, for example, smart phones, laptop computers, personal digital assistants, personal music players, or other personal devices.

Embodiments that operate using requests for access to a space are not limited to operating with any particular types of requests to access a space or information regarding such accesses. In some embodiments, a person may make a request to access a space by presenting a credential that identifies that the person is permitted to access to the space. The credential may be uniquely associated with the person. The credential may be in any suitable form, such as tangible object like a badge or card. The credential may be, for example, an identification badge for an employee. A request for access with such a tangible object may be made in any of various manners, as embodiments are not limited in this respect. The request for access may be made, for example, by using a reader to scan a bar code on the tangible object, swiping a magnetic strip on the tangible object through a reader, initiating near field communications (e.g., RFID) between the tangible object and a reader, or any other way of providing credential data from the tangible object to a reader.

Various examples of utilization analysis facilities are described in detail below. It should be appreciated, however, that embodiments are not limited to operating in accordance with any of these examples. Prior to a discussion of these examples, a discussion of the limitations the inventors have recognized in prior techniques is provided in connection with FIGs. 1A and IB. Some advantages of the various examples of utilization analysis facilities described below may be better understood in light of these deficiencies.

FIG. 1A illustrates a flowchart of a previously-used technique that attempted to approximate occupancy of an office building through identifying "attendance" at that office building on a specific day. The process 100 illustrated in FIG. 1A determines the attendance by relying on information regarding presentation by employees of the office building of credentials to devices that scan those credentials. Such credentials may be employee ID badges that are scanned using a bar code reader or RFID badge.

Specifically, the process 100 of FIG. 1A begins in block 102, in which a swipe of an employee badge by a person is detected by the system executing the process 100. The badge swipes used in this method may be at the front desk of the office building or in any other location in the office building, as the method treats all badge swipes in all areas of the building as identical regardless of location or time. In block 104, the system identifies the person performing the badge swipe detected in block 104 and determines whether a badge swipe by that person has been previously detected that day. If so, the system identifies that person as already "counted" for that day and returns to block 102 to detect another badge swipe. If, however, the system determines that the person performing the badge swipe detected in block 102 has not been previously detected that day, in block 106 the system increments a running total for the number of people in the office that day. Once the running total is incremented, the system returns to block 102.

FIG. IB is a graph illustrating the results of the method shown in FIG. 1A. The graph shows a bar graph of a number of badge swipes detected in blocks of time over the course of a day. In each block of time, a certain number of badge swipes are conducted, with the highest number per hour in the morning and around mid-day. The graph of FIG. IB also includes a dashed line indicating the result of processing that data using the technique of FIG. 1 A. In particular, the dashed line illustrates the total of the unique persons that performed the badge swipes, as determined by a system performing the process 100 of FIG. 1A.

The inventors recognized and appreciated three disadvantages of the technique illustrated in FIGs. 1A and IB that limit the utility of the information reported by that technique. First, the attendance for an overall space may not accurately reflect usage of portions of that space. For example, attendance may not inform a business of how many cubicle desks were used that day or how many personal offices were used that day, or whether any conference rooms were used and how those conference rooms were used. Some percentage of the people who arrived at work that day (and are counted in the attendance number) may be employees assigned to cubicles and some may be assigned to personal offices, but the technique cannot distinguish between them and thus cannot identify how the various desks in the office were used. Thus, the business cannot know whether it should have more or fewer cubicles, or more or fewer personal offices, based only on attendance. Moreover, a determination of attendance for an entire day as a whole may not indicate how any of the spaces within the office were used over the course of the day. The technique may not identify how people moved within the office during the day and how well the space suited those movements, such as whether conference rooms were or were not used during that day and how well the conference rooms matched the needs of the business.

Second, by treating all badge swipes, regardless of time, as identical, the process 100 effectively assumes that all employees arrived at the same time and left at the same time, and may not consider variations over the course of the day. This is not typically the case.

Employees arrive and depart throughout the day, and the number of employees within an office building varies over the course of a day. Thus, while the number of employees in the office may have been higher in the afternoon than in the morning, the technique of FIGs. 1 A and IB is unable to identify that fact. Further, the technique of FIGs. 1 A and IB is unable to indicate how the change in the number of employees affects usage of the office or spaces within the office over the course of the day. Third, the technique illustrated in FIGs. 1A and IB may not account for employees who move between buildings in a corporate campus with multiple buildings. Because that technique merely counts a number of unique badge swipes for a building, an employee who moves from a building where his/her workstation is located to another building for a short time may cause the technique to erroneously report attendance in both buildings. Specifically, the technique may "double -count" the employee by reporting the arrival of the employee at both buildings as an increase in attendance of both buildings. The employee obviously did not spend all day at both buildings (indeed, the arrival at the second building indicates that the employee left the first building), but this attendance technique will effectively report just that.

In sum, techniques that only count the employees that arrived at work one day may not be able to report how those employees interacted with spaces within the office over time. Those techniques are therefore unable to report key measures that may be most useful to a business seeking to evaluate how well its spaces accommodate its needs: an occupancy of the office or spaces within the office, and a utilization of the office or spaces within the office.

Occupancy and utilization may provide an organization with information on how its spaces are or are not used and therefore may aid the organization in determining whether its spaces are suitable. Attendance, as discussed above in connection with FIGs. 1A and IB, is a count of people who entered a space over a time period, without regard to whether or when the people left the space. Occupancy is a measure of a capacity of a space as compared to a number of people within that space at a time, such as a ratio of people (literal people or inferred people, as discussed below) within the space to the capacity of the space at a time. More specifically, occupancy is calculated based on an assignment of employees to spaces, such as the case where a particular employee is assigned to work in a particular cubicle or personal office. Occupancy of a space is therefore measured by a ratio of a number people that are assigned to the space and are currently present in the space to the capacity of that space.

Capacity may be a measure of a number of people a space is designed to accommodate, which may be a single person in the case of a personal office or six or eight people (or any other number of people) in the case of a conference room.

Utilization is related to occupancy, in that utilization is a measure of occupancy of a space during a period of time. However, whereas occupancy relates to a number of occupants assigned to use a space to the capacity of the space, utilization relates more generally to a number of occupants present in a space during a time interval to a capacity of the space. In a determination of utilization, the occupants that are considered may be occupants that are assigned to a space and/or occupants that are not assigned to the space. For example, in a business context, occupants may include people who are not employees, such as visitors to the business, or may be employees who are not assigned to a space.

In some embodiments described herein, utilization of various spaces may be determined using utilization analysis facilities. For example, a utilization analysis facility of some embodiments may determine utilization of a space (and sub-spaces within that space) from an analysis of electronic communication signals detected within the space. As another example, a utilization analysis facility of some embodiments may determine utilization of a space (and sub- spaces within that space) from an analysis of information indicating an arrival of a person to a space, which may be a request for access to a space.

As mentioned above, embodiments are not limited to evaluating the utilization of any particular space of any particular type. In some embodiments, utilization of a business' office space may be analyzed, but embodiments are not so limited. Further, it should be appreciated that embodiments may work with spaces that are arranged in any suitable manner, including spaces that include within them sub-spaces or that are themselves sub-spaces of a larger space.

For example, in some embodiments utilization may be analyzed for one, more, or all of various spaces arranged in a hierarchy of spaces that are related to one another. The spaces may be related to one another in any suitable manner, such as by being located within one building or being owned and/or occupied by one organization (e.g., business). The hierarchy of spaces may include multiple levels and may include multiple spaces at each level, with spaces at levels lower down in a hierarchy being sub- spaces of spaces at levels higher in the hierarchy.

One specific example of a hierarchy with which some embodiments may operate may be associated with a business that has multiple offices at geographically distributed locations (e.g., located around the world). The hierarchy, at the highest level, may include a conceptual space that includes all of the physical spaces operated by the business. The next level(s) down in the hierarchy may be conceptual spaces that are each associated with a particular geographic region (e.g., country or province) and include each of the physical spaces operated by the business in that geographic region. The next lower level of the hierarchy may include the campuses operated by the business in each of those geographic regions, where a campus may include one or more buildings. The next level in the hierarchy may include each of the buildings on a campus, and the next level may be each of the floors within those buildings. The next levels may be associated with each of the spaces within a floor, which may be organized into zones and individual rooms within zones.

It should be appreciated, though, that while one specific example of a hierarchy has been described, embodiments may operate with other hierarchies. For example, other embodiments may determine utilization for a business that operates an office that is only a single floor, and a hierarchy may be used that includes the office as a whole, zones of that office, and rooms within those zones. Embodiments are not limited to operating with any particular type of spaces or hierarchies of spaces.

FIG. 2 illustrates an example of a space with which some embodiments may operate. Some examples described further below may be described in the context of the example space of FIG. 2, but it should be appreciated that embodiments are not limited to operating with this example.

The space 200 of FIG. 2 includes five sub-spaces, labeled in FIG. 2 as cubicle area 202, personal office 204, personal office 206, conference room 208, and corridor 210, and storage 212. Space 200 also includes a computing device 214 executing instructions of a utilization analysis facility and storing in a data store 214A information regarding the space 200 and utilization thereof. (For ease of description, the utilization analysis facility executing on the computing device 214 may be referred to below as the utilization analysis facility 214.)

In some embodiments, spaces may be categorized into one of a set of categories that describe the function or usage of that space. Embodiments are not limited to operating with any particular set of categories. In some embodiments, the categories of spaces may be:

workstations, which are or include spaces at which individuals work, such as cubicle areas or offices; meeting spaces, which include spaces such as conference rooms; support spaces, which are not assigned to specific individuals but are often found alongside spaces used by individuals; private spaces, which are spaces that might otherwise be included in another category but that are to be left unanalyzed due to personal privacy concerns (e.g., bathroom spaces); and excluded spaces, which are spaces that might otherwise be included in another category but that are to be left unanalyzed (or analyzed but unreported) due to any other business reason (e.g., an executive office for which utilization is not reported).

Accordingly, referring again to FIG. 2, the space 200 may include spaces of various categories. Cubicle area 202 may be treated as a whole as a shared workstation area, or may be sub-divided into multiple different individual workstation areas. Personal office 204 may be an individual workstation area and may be an assigned workstation, whereas personal office 206 may be an unassigned, shared workstation area. Conference room 208 may be a meeting space, and corridor 210 and storage 212 may both be designated as unoccupied support spaces.

Spaces may be associated with a capacity, which is a maximum number of individuals the space is intended to hold at a time. For example, personal offices 204, 206 may each have a capacity of one, conference room 208 may have a capacity of six, and cubicle area 202 may have a capacity of four. Spaces that are intended to be unoccupied, such as corridor 210 and storage 212, may in some cases not be associated with a capacity. A capacity of a space may in some cases be expressed as a number of seats in an area. For example, personal office 204 may have one seat and thus have a capacity of one. For ease of illustration, seats are not illustrated in FIG. 2.

Using techniques described herein, the utilization analysis facility 214 may determine utilization of the space 200. The facility 214 may do so in part by determining utilization of each of the sub-spaces 202-212 of the space 200. The facility 214 may, for example, analyze information indicative of usage of each of the spaces 202-212 and infer a number of occupants in each space during time intervals, and compare that number of occupants to the capacity of each of the spaces to determine utilization of the spaces. Further, as discussed below, the facility 214 may determine capacity information for the space 200 as a function of the capacities of each of the sub-spaces 202-212. For example, the facility 214 may determine the capacity of the space 200 as a sum of the capacities of spaces 202-212, or as a sum of the capacities of spaces in some specified categories of space based on user input of specific space categories to include or exclude. Once capacity of the space 200 is determined, the facility 214 may also determine utilization of the space 200, by inferring a number of occupants in the space 200 in time intervals.

Various techniques are described herein for determining utilization of a space and embodiments are not limited to determining utilization using any particular one of these techniques or combination of these techniques. For example, the utilization analysis facility 214 may determine utilization information from processing information regarding electronic communication signals emitted by devices in the spaces 200-212. As another example, the utilization analysis facility 214 may determine utilization from processing information that is suggestive of the presence of a person, such as information that indicates the arrival of a person, as well as inferences drawn by the facility 214 regarding departures of occupants from a space. Such information indicative of the presence of an occupant in a space may include information indicating usage of a network-connected device in the space, such as usage of a hardwired device such as a computer or Voice-over-Internet-Protocol (VoIP) phone that is connected to a wireline network port in the space. Information indicative of the presence of an occupant may also include physical occupancy sensors, which may include sensors installed in desks or in seats to detect the presence of an individual. Information indicative of the presence of an occupant may also include cameras and image processing engines that are configured to detect the movements of occupants, such as cameras installed above or near entrances to spaces. Presence may also be indicated by requests for access for a space, such as requests for physical access or requests for electronic access. Where such information is used, utilization analysis facility may identify an arrival of an occupant to a space and make a prediction of a length of time that the occupant will spend in the space. The facility may then use the prediction of the length of time that the occupant will spend in the space to identify a time that the occupant will depart the space, and use such information on arrivals and departures to determine utilization of the space.

Specific examples of techniques the facility 214 may use to make these determinations are described in greater detail below. To provide context for those discussions, FIG. 2 illustrates some of the sources of information that may be used by the facility 214.

For example, the facility 214 may determine utilization from analysis electronic communication signals emitted by devices 220-232. Network infrastructure devices 220-220C may both emit electronic communication signals and detect such signals. The devices 220-220C may also process the signals and information regarding those signals may be stored in data store 214A for analysis by facility 214.

The signals that are processed by the facility 214 may include signals collected over a period of time that is longer than the time intervals by which utilization may be determined. For example, while the time intervals by which the facility 214 may determine utilization of spaces may be less than a day, the signals processed by the facility 214 may to determine that information may have been collected over a period of time longer than a day. As specific examples, the time interval for utilization determinations may be a fraction of an hour (e.g., a half hour or 15 minutes) while the signals that are analyzed may have been collected over a period of time longer than a week or a month.

In some embodiments, the signals collected over the longer period of time may be used to train machine learning techniques employed by the facility 214, such as filtering or analysis techniques, after which the trained techniques may be applied to signals in each time interval to determine utilization information.

Any suitable information regarding signals may be stored in the data store 214A for use by the facility 214. For example, the information may include, for each detected signal, a time that signal was detected by a device 220-220C, a received signal strength (RSS) for that signal, an identifier for the device that emitted the signal, and an estimated position in the space 200 of the device that emitted the signal, among other information. The identifier for the device may be any suitable identifier, as embodiments are not limited in this respect. In some embodiments, the identifier may be a hardware identifier that identifies hardware of the device, such as the device itself or a component of the device. For example, the hardware identifier may be a Media Access Control (MAC) address.

The facility 214 may analyze the information stored in data store 214A in any suitable manner to determine utilization information for the space 200. For example, the facility 214 may analyze the signal information so as to identify a limited set of devices that may be indicative of utilization of spaces. The limited set of devices may be those that the facility 214 infers are operated by individual occupants of the space 200 and are uniquely associated with individual occupants among the devices 222-232.

As used below unless indicated otherwise, an "occupant" is an inferred individual that the facility 214 identifies from an analysis of information suggestive of utilization. The facility 214 may be designed to identify individual persons and, in a best case, each occupant identified by the facility 214 may precisely correspond to one individual that is present in the space 200. However, due to uncertainties of algorithmic analysis resulting from limited information, at a time the facility 214 may identify two occupants that are, in fact, the same individual person. Though, as the facility 214 processes more information, the facility 214 may identify the mistake and merge the two occupants into one occupant.

In some embodiments, it may be important to maintain the privacy of individuals in a space and techniques may be used to anonymize information about the individuals that is used in tracking occupants and determining utilization. For example, information that identifies an occupant or a device may be anonymized such that while movements of unique occupants can be tracked over time to determine utilization, those occupants are not matched to specific individuals. In such embodiments, no identifying information for specific individuals may be associated with the occupants. Any suitable anonymizing techniques may be used, as embodiments are not limited in this respect. In some embodiments, one or more hash functions may be used to process unique identifiers for individuals or devices to produce corresponding identifiers that, while uniquely corresponding to those individuals/devices, cannot be used to identify the individuals/devices and therefore maintain the anonymity of the individuals/devices.

The facility 214 may apply various complex statistical analysis and machine learning techniques to identify the limited set of devices. For example, the facility 214 may analyze the electronic communication signals detected in the space to identify devices that are located outside of the space 200, such as mobile phone 232. The facility 214 may also analyze the electronic communication signals to identify network infrastructure devices, such as devices 220- 220C. The facility 214 may also analyze the electronic communication signals to identify multiuser equipment, such as printer 222, and to identify stationary devices, such as desktop computers 224. By identifying such network infrastructure, multi-user equipment, and stationary devices and devices outside the space 200, the facility 214 may filter from its analysis signals emitted by those devices and instead analyze those signals emitted by devices that are inside the space 200 and that move with occupants in the space 200, which are illustrated in FIG. 2 as devices 226-230E. Devices 226-230E, as illustrated in FIG. 2, include laptop personal computers and mobile phones, which a person may carry as they move about space 200. By analyzing signals emitted by devices 226-230E, the facility 214 may identify signals emitted by devices that are indicative of utilization of the space 200. The facility 214 may also analyze the electronic communication signals emitted by those devices 226-230E to identify, from among those devices, ones that are uniquely associated with individuals present in the space 200. The facility 214 may do so by inferring the existence of occupants operating each of the devices, and identifying devices that appear to be operated by the same occupant. For example, from among the devices 226-230E, the facility 214 may infer a relationship between devices and occupants, so as to identify devices that are operated by the same occupant, and may identify that mobile phone 226 and laptop computer 228 are operated by a single occupant. The facility 214 may then select one of those devices 226, 228 to filter from its analysis such that signals emitted by the filtered device are no longer considered by the facility 214 in determining utilization. By doing so, the facility 214 may identify a limited set of devices (e.g., 226 and 230A-230E) with a one-to-one correspondence between devices and occupants inferred to be present in the space 200. Having identified a set of devices that uniquely indicate the presence and movements of occupants, the facility 214 may then analyze the signals emitted by the devices of the limited set to determine utilization of the space 200.

Various examples of ways in which the facility 214 may perform these analyses of electronic communication signals are described below in connection with FIGs. 3-11.

While in some embodiments the facility 214 may evaluate electronic communication signals to determine utilization information, in other embodiments the facility 214 may additionally or alternatively evaluate information indicative of the presence of an occupant in a space to determine utilization of a space. More specifically, in some embodiments, the facility 214 may evaluate the information indicative of presence of occupants together with inferred departure times for those occupants from those spaces to determine utilization. The information indicative of presence of an occupant may, in some embodiments, be information indicating arrival of an occupant to a space. An arrival of an occupant to a space may be indicated, for example, by a request for access made by an occupant, which may be a request by the occupant for physical access to a space.

A request for physical access may be made in some embodiments with a swipe/scan of a credential, such as an employee identification badge, a biometric identifier (e.g., fingerprint scan or retinal scan), or any other suitable credential. While not illustrated in FIG. 2, a credential authentication device may be located at the entrance of the space 200 and access to the space 200 may be restricted to those having credentials that indicate permission to access the space 200.

In some embodiments, data indicative of presence of an occupant in a space, such as information regarding requests for access made to a credential authentication device, may be stored in the data store 214A for analysis by the utilization analysis facility 214. For example, for each request for access to a space, the data store 214A may store information on a time the request for access was made, information identifying a person associated with a credential (if available), information identifying the space for which access was requested, and information regarding the credential.

The facility 214 may analyze the information stored in data store 214A in any suitable manner to determine utilization information for the space 200. For example, in some embodiments the facility 214 may analyze each piece of information in the data store 214A that is indicative of presence of an occupant in a space and determine a corresponding inference about when that occupant will leave that space. Based on both the information indicative of presence and the inferred departures for the occupants, the facility 214 may determine utilization for the space 200. Thus, for example, the facility 214 may receive a set of requests for access to a space and, based on those requests, infer a set of departure times and subsequently determine utilization based on the set of requests and set of inferred departures. Though, in other embodiments the data store 214A may store information that explicitly indicates departures, which may be the case where individuals are required to present credentials again when leaving a space. In these other embodiments in which the data store 214A stores sets of requests to depart a space, the facility 214 may determine utilization from an analysis of the sets of requests for access to a space and sets of requests to departure the space.

Various examples of ways in which the facility 214 may perform these analyses of information indicating the presence of occupants, including for inferring departures from spaces based on an analysis of the information, are described below in connection with FIGs. 12-15.

FIG. 3 illustrates a process 300 that may be implemented by a utilization analysis facility in some embodiments for determining utilization of a space. The process 300 of FIG. 3 may be used in embodiments that base determinations of utilization at least in part on an analysis of electronic communication signals detected in a space. While the example of FIG. 3 may be described in the context of the space 200 illustrated in FIG. 2, it should be appreciated that the technique of FIG. 3 is not limited to operating with the space 200 or spaces similar to the space 200.

Prior to the start of the process 300, an organization may be operating in a space, such as a business operating in an office. Various parts of that space may be allocated to different purposes, as was the case with spaces 202-212 of the space 200 of FIG. 2. For example, some parts of the office may be allocated as personal offices, and some of those personal offices may be assigned for use to specific individuals associated with the organization (e.g., employees). The organization may desire information on how its space is being used, including whether its space is being effectively used. To make that determination, the organization may operate a utilization analysis facility to analyze information suggestive of usage of the space and report on utilization of the space. The utilization analysis facility may, in some cases, operate within the space, such as was the case of FIG. 2 with facility 214 operating within the space 200. In other embodiments, however, information about usage may be collected from the space and then analyzed by a utilization analysis facility located outside the space, as embodiments are not limited in this respect.

The process 300 begins in block 302, in which the utilization analysis facility receives as input a floor plan of the space for which utilization is to be analyzed, and initializes a coordinate system of that floor plan and of reported locations of signals. The floor plan may be used by the facility to localize devices within the space to be analyzed based on the signals emitted by the devices. By localizing the devices within the space, specific sub-spaces within which the devices are operating can be identified, which can aid the utilization analysis facility in identifying the utilization of sub-spaces of the space.

With respect to the coordinate system for reported locations of signals, in some embodiments the utilization analysis facility may receive as input regarding signals information that indicates an estimated location of the device that emitted that signal. The estimated location may be reported in a two- or three-dimensional coordinate system. Such information may be received, for example, when the signals are analyzed by a device that detects the signals and localizes devices within a coordinate system with which it is configured. Such a localizing device may, in some cases, be network infrastructure devices such as wireless access points or access points paired with analysis engines. Some commercial wireless access points, in addition to performing typical networking operations, may be configurable to operate with analysis engines to estimate the locations of devices that emit signals received by the wireless access points. For example, the "Connected Mobile Experiences" (CMX) product line, available from Cisco Systems, Inc., may operate with wireless access points to report estimated locations of devices that emit signals received by the wireless access points. Aruba Networks, Inc., similarly offers an Analytics and Location Engine (ALE) that may process signals received by a wireless access point to report estimated locations. Navizon, Inc.'s Indoor Triangulation System (ITS) and Orb Analytics, Inc.'s Non- Invasive Location Analytics (NILA) system similarly operate to estimate locations of devices based on detected signals. It should be appreciated that embodiments that receive an estimated location as input are not limited to operating with any of these examples of localizing systems, but instead may operate with any system that outputs an estimated location of a device in a coordinate system. In some cases, the coordinate system of the localizing device may not align with a coordinate system for a floor plan, and the initialization may be performed to enable a translation between the coordinate systems. The initialization of the coordinate system of the floor plan and of the coordinate system for the reported locations for signals may be carried out in any suitable manner, including using known techniques for normalizing coordinates to a common coordinate system. For example, in some embodiments a geographic coordinate system of latitude and longitude may be used as a normalized coordinate system. Using known techniques, the floor plan of the space may be mapped to geographic (e.g., latitude and longitude, and in some cases altitude) coordinates, such that the space and sub-spaces within the space are each defined by a polygon of geographic coordinates. Similarly, using known techniques the coordinate system by which estimated locations of devices are reported may be mapped to geographic coordinates. By normalizing the coordinate systems, the utilization analysis facility may be able to transform input coordinates relating to the estimated position of a device to the normalized coordinate system, then identify a specific location of that device within the space being monitored.

The initialization of the coordinates systems in block 302 may be performed in any suitable manner, including through one or more interactions between the utilization analysis facility and a user via a user interface, in which the user provides input to perform the initialization.

Once the initialization is performed, the utilization analysis facility may begin analyzing information regarding electronic communication signals detected in the space. Accordingly, in block 304 the utilization analysis facility receives data indicating electronic communication signals that were detected in the space to be analyzed. The signals identified in the data may be signals collected over a period of time that is long as compared to a period of time for which utilization is to be determined. For example, in some embodiments the utilization analysis facility may determine utilization information using signals detected in a space over the course of a first period of time, which may be a day, three days, or any other amount of time. To determine such utilization information, however, the utilization analysis facility may be trained using information regarding signals detected in the space over a second period of time that is different from and potentially longer than the first period, such as a week, a month, or more. In some embodiments, the second period of time may include the first period of time, though embodiments are not so limited.

The information received in block 304 may be in any suitable form, including in a list of detected signals and information regarding those detected signals. The information that is received may be information output by a localizing process, which may be any suitable process (e.g., a triangulation process performed by multiple wireless access points) including one carried out by one of the commercial localizing engines discussed above, and may therefore include for each signal information identifying an estimated position of a device that emitted that signal. The information for each signal may also include a time that signal was detected, a received signal strength (RSS) for that signal, and an identifier for the device that emitted the signal, among other information. The identifier for the device that emitted each signal may have been determined in any suitable manner, including by extracting the identifier from the signal itself. For example, a MAC address may be extracted from a signal and used as the identifier for the device that emitted that signal. In some embodiments, the information received in block 304 for a signal may include information on content of the signal, such as a protocol used in

communicating the signal.

As mentioned above, the electromagnetic spectrum of a space may be chaotic, including signals emitted by devices inside and outside the space and with signals of varying quality in the data. Accordingly, as a first step of analyzing the information regarding the signals to determine utilization of a space, the utilization analysis facility may in block 306 filter and clean the data to remove erroneous or irrelevant data, where irrelevant data may relate to signals that are not indicative of utilization of a space. Signals that are not indicative of utilization may be signals from devices located outside the space to be analyzed. Any suitable process may be used in block 306 to filter and clean the data, as embodiments are not limited in this respect. Particular techniques that may be used in some embodiments for filtering and cleaning the data are discussed below in connection with FIG. 10.

Once the data received in block 304 is filtered and cleaned in block 306, the utilization analysis facility may begin being trained using the data resulting from that filtering/cleaning process. The training may be performed in blocks 308 and 310, and may be performed to train the device based on the signals emitted in the space over time to identify those devices in the space that are associated with and operated by individual occupants of the space. By identifying the devices that are associated with and operated by individual occupants, the utilization analysis facility may identify the signals emitted by those devices and use those signals to determine utilization information.

Specifically, in block 308 the utilization analysis facility may analyze the information regarding signals received in block 304 to identify, based on the signals, devices that are not operated by and/or carried by individual occupants. Such devices may include network infrastructure devices, multi-user equipment, stationary devices, or other devices that are not operated by individuals and/or are not carried by individuals. Such devices may emit electronic communication signals that may be detected in the space, but may not be highly indicative of utilization of a space. Accordingly, signals emitted by such devices may be filtered from analysis.

Of the remaining devices indicated by the information received in block 304 (i.e., the devices identified as operated by and/or carried by occupants), in block 310 the utilization analysis facility may identify from among those devices a limited set that are uniquely associated with occupants.

As discussed above, an occupant may be an inferred individual that is algorithmically identified by the utilization analysis facility, such as from an analysis of electronic

communication signals. Occupants may be intended to correspond one-to-one to actual people. However, those skilled in the art will appreciate that due to limitations and uncertainties of algorithmic analysis, including potential limitations on the data available for training, a utilization analysis facility may inadvertently identify one individual as two occupants.

By identifying devices that are, among the devices left following the filtering of block 308, shared by occupants and subsequently identifying a set of devices that are, among those devices, uniquely associated with occupants, the utilization analysis facility may identify a set of devices that are each indicative of a single occupant. By tracking locations and movements of those devices over time, as determined from signals emitted by those devices, the utilization analysis facility may determine utilization of a space.

Illustrative techniques for performing the training of blocks 308, 310 are discussed in more detail below in connection with FIGs. 4-6.

In block 312, once the utilization analysis facility determines the limited set of devices that are uniquely associated with occupants of a space, the facility may track movements of occupants in the space during a time period of interest. More specifically, by analyzing signals (identified in the information received in block 304) emitted by devices in the limited set and detected during the time period of interest, the utilization analysis facility may identify movements of those occupants during the time period of interests. The facility may, for example, identify sub-spaces in which the occupants were located during time intervals of the time period of interest. Thus, if the time period of interest is a day, the facility may analyze signals emitted from the devices of the limited set to identify, for each time interval within the day (e.g., each half hour interval, each 15 minute interval, or any other suitable interval) whether the occupant is present within the space and if the occupant is present within a particular sub- space of the space.

From this movement information collected in block 312, in block 314 the utilization analysis facility may determine utilization information for the space during the time period of interest. Examples of ways in which the analysis of blocks 312, 314 may be carried out are discussed in detail below in connection with FIGs. 7-8. As part of determining utilization, a facility (using any of the techniques described herein) may determine a utilization of a space over a particular time period, and may additionally determine a number of occupants in the space during the time period. The facility may also determine any relevant statistics regarding the utilization or number of occupants in the space during a time interval, which the facility may determine through analyzing utilization over a period of time longer than the time interval. For example, the facility may determine peak utilization or number of occupants during a day or week or an average utilization or number of occupants during a day or week by analyzing the utilization and number of occupants in that space over the course of a longer time period (e.g., a week or month).

Once the facility determines the utilization information for the space, the process 300 ends. Following the process 300, the information on utilization determined by the utilization analysis facility may be output to a user in any suitable manner. For example, in some embodiments, the utilization information may be output to a user in a form that enables the user to view utilization for the space as a whole, or for specific sub-spaces or combinations of sub- spaces within the space. Further, in some embodiments, the utilization analysis facility may further process the information regarding utilization to make recommendations to a user on how to increase utilization. Examples of techniques that a utilization analysis facility may use to make such recommendations are discussed in greater detail below in connection with FIGs. 16- 17.

As discussed above, in some embodiments a utilization analysis facility may identify devices that are not operated by or associated with individuals, or carried by individuals. Such devices may include network infrastructure devices, multi-user equipment, and stationary devices. In these embodiments, such devices may not contribute much information by which a utilization analysis facility could determine utilization of a space, but the signals detected in a space may include signals emitted by these devices. To ease determination of utilization information from analysis of electronic communication signals, these devices may be identified such that signals emitted by these devices may be excluded from analysis.

A utilization analysis facility may carry out any suitable process for identifying devices to exclude from analysis (also known as "blacklisting" devices), as embodiments are not limited in this respect. FIG. 4 illustrates a process 400 that may be used in some embodiments to determine devices to exclude.

Prior to the start of process 400, electronic communication signals may have been detected in a space and processed by a localizing engine to produce an estimated position of a device that emitted each signal. A utilization analysis facility performing the process 400 may therefore analyze information regarding each signal that includes at least an estimated location for each signal, as well as information regarding the device that emitted the signal, such as an identifier for the device. The identifier may be any suitable identifier, including a fixed hardware identifier for the device. Such a fixed hardware identifier may be, for example, a MAC address for the device that emitted the signal, which may have been included in the signal that was detected. The information about the device that emitted each signal may be useful to identify signals in the listing that were emitted by the same device, such that actions taken by that device over time can be analyzed. The information regarding each signal may additionally include time information that indicates a time at which each signal was detected in the space.

The process 400 begins in block 402, in which a utilization analysis facility analyzes each of the signals identified in a set of information about signals detected in a period of time. The facility analyzes the signals (or information about the signals) in block 402 to attempt to infer a type of device that emitted each of the signals. The facility may analyze the signals in any suitable manner. In some embodiments, the facility may analyze content of the signals.

For example, in some cases a signal may include an identifier for a device that emitted the signal, and the facility may analyze those identifiers to determine whether each identifier is indicative of the type of the device. MAC addresses, for example, are formed from a concatenation of two numbers: one an identifier for a manufacturer or vendor, and another an identifier for a specific device sold by that manufacturer or vendor. In some cases, a manufacturer or vendor may sell only devices of a certain type, primarily sell devices of a certain type, or have a substantial portion of the number of devices sold be of a certain type. For example, Cisco Systems, Inc. primarily sells network infrastructure equipment, and there is therefore a strong likelihood that a device having a MAC address for which the first portion is the identifier for Cisco Systems should be categorized as a piece of network infrastructure equipment. Similarly, Lexmark primarily sells printers and other office equipment, and there is therefore a strong likelihood that a device having a MAC address for which the first portion is the identifier for Lexmark should be categorized as a piece of multi-user equipment. Though, in other cases a MAC address or other hardware identifier included in a signal may not be dispositive (or likely to be dispositive) of a categorization of a device. For example, a MAC address for which the first portion indicates Apple, Inc. may not be dispositive of the type of device, as Apple, Inc. sells laptop and desktop personal computers, mobile phones, and network infrastructure equipment, among other products. Thus, in embodiments that review a hardware identifier, the utilization analysis facility may analyze the hardware identifier to determine whether it is indicative of a specific categorization of device. As another example of analysis that may be carried out in block 402, in some embodiments the facility may identifier a type of data communicated in each signal emitted by a device to identify the categorization of a device. In such embodiments, the facility may determine whether a type of data or a protocol used in the signal is indicative of a particular categorization of device. For example, a device that only emits signals in accordance with the Internet Printing Protocol (IPP) may be likely to be a printer, and therefore be a piece of multiuser equipment. Any other analysis may be performed on signals to identify a type of the signals, as embodiments are not limited in this respect.

As another example of a way in which the utilization analysis facility may perform the analysis of block 402, in some embodiments a statistical analysis and/or machine learning algorithm may be used to identify a type of device based on the signals. For example, using a set of devices that have been manually labeled with particular types and a set of signals exhibited by those labeled devices, a classifier may be trained for each type of device with characteristics of signals emitted by devices of that type. Such characteristics may include any suitable information that may be inferred from the signals, such as timing or location characteristics that may relate to times at which the devices emit signals, locations at which the devices emit signals, movements of the devices as exhibited by the signals, protocols used by the devices, etc. Once the classifiers are trained using the labeled data, the machine learning algorithm may be used to determine a type of device based on signals emitted by that device. More specifically, the utilization analysis facility may review the signals emitted by that device using the machine learning algorithm to determine which classifier (and corresponding classification) most closely matches that device. Those skilled in the art will appreciate how to implement a machine learning algorithm to carry out such an analysis.

In block 404, the utilization analysis facility may also analyze the information about the signals to identify movements, if any, of the device over time. The information analyzed in block 404 may include the estimated locations of the device that emitted each of the signals. By analyzing the estimated locations for each of multiple signals emitted by a particular device over time, the utilization analysis facility may determine how that device moves over time.

Movement or lack of movement of a device may be indicative of how that device should be categorized. For example, if the signals indicate that the device never moves, the device may be more likely to be a piece of networking infrastructure equipment or multi-user equipment, or likely to be a stationary device, and may therefore be categorized as such.

Similarly, in block 406 the utilization analysis facility may analyze the signal emission times for each signal emitted by a device to determine whether the signal emission times are indicative of how the device should be categorized. For example, a device that emits signals 24 hours a day indefinitely may be more likely to be a network infrastructure device than a device associated with an occupant. Conversely, a device that emits signals only between 9 and 5 on weekdays may be more likely to be associated with an occupant.

Once the analyses of blocks 402-406 have been performed, the utility analysis facility may determine whether a categorization can be determined for each device that emitted signals that were analyzed. The utility analysis facility may determine whether the device can be categorized in any suitable manner. In some embodiments, for example, a score may be calculated for each device through performing a weighted sum of factors produced from each of the analyses of blocks 402, 404, 406 and analyzed to determine whether the weighted sum is indicative of one of the blacklisted categories: infrastructure devices, stationary devices, or multi-user equipment. In other embodiments, a machine learning process may be employed that may be trained with characteristics exhibited with devices properly categorized into each of those three categories, after which the machine learning process may be used to determine whether the characteristics of devices determined from the analyses of blocks 402, 404, 406 can be used to categorize the devices into one of the blacklisted categories. Any other suitable technique for categorizing devices based on the analyses of blocks 402-406 may be used, as embodiments are not limited in this respect.

The utilization analysis facility may then, in blocks 408-416, loop through each of the devices indicated by the signals analyzed by the facility and determine whether to blacklist those devices. Specifically, in blocks 408-412, the utilization analysis facility may determine whether a device is categorized (as a result of the prior analysis) as an infrastructure device, a stationary device, or multi-user equipment. If categorized into any of those categories, in block 414 the utilization analysis facility filters (what may be termed "blacklists") that device from further analysis. In block 416, the facility determines whether the loop should continue with evaluating more devices in the list and, if so, performs the acts of blocks 408-414 for another device. If, however, each of the devices has been analyzed, the process 400 ends.

Following the process 400, the utilization analysis facility may have identified a set of devices that transmitted signals detected in the space and that appear from the analysis of process 400 to be devices that are operated by occupants of the space. Such devices may be more indicative of utilization than the blacklisted devices, and signals emitted by those devices may therefore be further analyzed to determine utilization. As mentioned above, in some embodiments steps may be taken to preserve anonymity of individuals when signals emitted by devices are analyzed to determine utilization.

Though, it should be appreciated that individuals using a space may carry with them two or more devices at a time. For example, an occupant may carry with him or her a laptop computer and a mobile phone. If a utilization analysis facility were to analyze each of those two devices as an independent indicator of utilization without accounting for the fact that the two devices are shared by one occupant, then the utilization reported by the facility could be incorrect: it could report the presence of two occupants in a space based on the two devices, whereas there was actually only one person in the space. Accordingly, in some embodiments a utilization analysis facility may additionally analyze signals emitted by the devices that were not filtered from a blacklisting technique (e.g., the technique illustrated in FIG. 4) to further filter that set of devices. Specifically, the utilization analysis facility may review signals emitted by devices in that set to infer devices that are shared by one occupant, and subsequently filter from analysis all but one device for each occupant.

FIGs. 5-6 illustrate examples of techniques that may be used in some embodiments to identify devices that are likely to be shared by occupants and to select a single device for each occupant for which signals will be evaluated in determining utilization of a space, which may be a device that the occupant uses as his/her primary device. It should be appreciated that embodiments are not limited to implementing the techniques illustrated in FIGs. 5-6.

Prior to the start of the process 500 of FIG. 5, a set of devices that are personally operated by occupants of a space is identified. That set of devices may have been identified in any suitable manner, including by detecting electronic communication signals in a space, identifying a set of devices that emitted those signals, and filtering from that set devices that are not devices personally operated by occupants (e.g., infrastructure devices, multi-user equipment, stationary devices, etc.). For example, the set of devices may have been identified through a filtering process similar to the one discussed above in connection with FIG. 4. The set of devices so identified may include devices that are shared by an occupant. In addition, prior to the start of the process 500, a set of information on signals emitted by the devices over a period of time is collected. This information may be of a same type of information as discussed in connection with block 304 of FIG. 3. The information may indicate multiple signals emitted by each device during the period of time, which a utilization analysis facility may use (as discussed below) to determine behaviors of the devices over time. In some embodiments, the information on the signals may have been filtered and cleaned, as discussed briefly above in connection with FIG. 3 and as discussed in more detail below in connection with FIG. 10.

The process 500 begins in block 502, in which a utilization analysis facility starts with an assumption that each device in the set of personal devices is operated by a different, unique occupant. The facility may act on this assumption by, for example, storing data identifying each device as having a different occupant. After identifying each device as having its own occupant, the utilization analysis facility may begin analyzing the information on the signals emitted by each of the devices to identify behaviors exhibited by each of the devices. Such behaviors may be exhibited through times and locations at which signals were emitted by the devices. The utilization analysis facility may identify devices that have similar behaviors as devices that are likely to be operated by a same occupant. This is because when the utilization analysis facility observes from the signals that two or more devices repeatedly transmit signals at the same times and same locations (or within a threshold proximity in time and location to one another), this may be indicative that the devices may have been carried to the locations by the same occupant and are therefore operated by the same occupant. Similarity in behaviors may be determined using any suitable technique, including known machine learning and clustering algorithms or, as discussed below, techniques for evaluating a Euclidean distance between devices in a multi-dimensional space.

Specifically, the utilization analysis facility analyzes the signals to identify the behaviors in blocks 504 and 506. In block 504, the facility reviews information on signals emitted by each of the devices in the set to identify times at which each of the signals were emitted. From an analysis of the times, the facility may identify characteristics of the timing of signal emissions for each device. Timing characteristics may include information describing times at which a device emits signals, such as times at which the device repeatedly or commonly emits signals. Examples of timing characteristics that may be determined include start and stop times for emissions of signal from a device for a day or a set of days, times of day when the device is repeatedly or commonly active and times of day when the device is repeatedly/commonly not active, or other information regarding times at which a device emits signals. In block 506, the utilization analysis facility reviews information on signals emitted by each of the devices in the set to identify locations of the devices at which each of the signals were emitted. From an analysis of the locations, the facility may identify characteristics of the locations of the signal emissions for each device. Location characteristics may include information describing locations at which a device emits signals, such as locations at which the device repeatedly or commonly emits signals. Examples of location characteristics that may be determined may include start and stop locations for emissions of signals from a device for a day, locations at which the device is repeatedly or commonly active or not active, or other information regarding locations at which a device emits signals. A path followed by a device during a period of time (e.g., a day) may also be determined from a set of locations at which emissions of signals from the device were detected. In embodiments, information regarding locations may be expressed as a set of two- or three-dimensional coordinates according to any suitable coordinate system, including a geographic coordinate system (e.g., latitude and longitude, and in some cases altitude as well). In blocks 508 and 510, the utilization analysis facility analyzes the behavior information collected in blocks 504, 506 to identify those devices having correlated behaviors and that appear likely to be operated by a same occupant. As discussed above, the analysis of the behaviors and identification of shared devices may be performed in any suitable manner. In some

embodiments, machine learning and clustering techniques may be used. Those skilled in the art will appreciate how to use clustering techniques to identify devices exhibiting similar behaviors. In other embodiments, an analysis of Euclidean distances may be used to make the

determination. An example using Euclidean distances is discussed in connection with blocks 508 and 510 below, but it should be appreciated that embodiments are not limited to operating with any specific technique.

In block 508, the utilization analysis facility compares usage times and movements for devices in the set (as determined in blocks 506, 508) to determine whether any devices have times/movements that indicate correlations between devices that may be suggestive of being shared by a same occupant.

In some embodiments, in block 508 the utilization analysis facility may identify every possible pair of devices in the set and, for each pair, calculate a Euclidean distance between the two devices based on an analysis of the timing characteristics and location characteristics for those two devices. For example, in some embodiments the timing characteristics may be start and stop times for signals emitted by a device during a day (i.e., the time a first signal of a day was emitted and the time a last signal of a day was emitted, as indicated by the signal information) and location characteristics may be start and stop locations for signals during a day (i.e., the locations from which a first signal and a last signal of a day were emitted). The location characteristics may be expressed as two-dimensional geographic coordinates. Each device in each pair may therefore be characterized using six values: start time, stop time, start x-coordinate (e.g., longitude), start y-coordinate (e.g., latitude), stop x-coordinate, and stop y-coordinate. A six-dimensional coordinate system may be constructed and the utilization analysis facility may calculate a Euclidean distance between the six-value points associated with the two devices in each pair of devices. Those skilled in the art will understand how to calculate a Euclidean distance between points in a six-dimensional coordinate system. In these embodiments, the distances are values indicative of correlations between the two devices.

Once the comparison of block 508 is performed and correlations identified, in block 510 the utilization analysis facility evaluates the correlations to identify devices that appear to be highly correlated and that appear to be evaluated by the same occupant. Such an evaluation may be carried out in any suitable manner, as embodiments are not limited in this respect. For example, in some embodiments, the utilization analysis facility may compare the distances to a threshold distance and identify pairs of devices having a distance below that threshold as devices that are operated by the same occupant.

Once the shared devices are identified in block 510, the process 500 ends.

Those skilled in the art will appreciate that the technique described above in connection with blocks 506, 508 for evaluating pairs of devices will enable the utilization analysis facility to identify two devices that are operated by a same occupant. In some environments, this may be sufficient as it may be unlikely that an individual will carry more than two devices (e.g., a laptop computer and a mobile phone). However, in other scenarios, it may be more likely for an individual to carry three devices (e.g., a laptop computer, a mobile phone, and a tablet computer), or any other number of devices. In such scenarios, a modification to the process 500 may be made. Accordingly, in some embodiments a user of the utilization analysis facility may provide as input a maximum number of shared devices per occupant. In some such

embodiments, the utilization analysis facility may continue evaluating potential shared devices and assigning devices to an occupant (where the technique indicates that the devices are shared by the occupant) until no more devices have correlations that indicate sharing or until an occupant is associated with the maximum number of devices. Accordingly, in some

embodiments the process 500 may additionally include a loop that continues executing blocks 506, 508 until no more devices have correlations suggestive of being shared by an occupant.

In some embodiments that enable groups of three or more shared devices to be determined, an additional step may be performed to aid in the identification of clusters of three or more devices. When a pair of devices has been identified as correlated and therefore likely to be operated by a same occupant, the system may merge the timing and location characteristics for those two devices to create merged characteristics for a cluster of devices associated with that occupant. The merging may be done in any suitable manner, including by averaging each of the individual pieces of information in the timing and location characteristics (e.g., the six values discussed above). Subsequently, to determine whether a third device is correlated with those two and operated by the same occupant, the timing and location characteristics for the third device may be compared (e.g., using the Euclidean distance technique described above) to the merged timing and location characteristics for the cluster. In some embodiments that implement such merging, at each iteration of a loop through blocks 508, 510, each of the pairs of devices (e.g., a pair of actual devices, or a pair of an actual device and a "merged device" that is a cluster) that the utilization analysis facility determines to be correlated may be merged, and the merged devices/characteristics may be used in the subsequent iterations of the loop. In some embodiments, the utilization analysis facility may merge devices into clusters unless a cluster includes a maximum number of devices that may be set by a user. Following the process 500, clusters of devices that are associated with individual occupants may be identified, where those clusters may include two or more devices.

Additionally, in some cases a number of devices may not have been clustered together and may each be associated with an occupant.

As discussed above, in some embodiments a utilization analysis facility may attempt to identify a set of devices that are uniquely associated with individuals in a space by identifying a set of devices that are uniquely associated with occupants of a space. Accordingly, once a utilization analysis facility identifies relationships between devices and occupants, including clusters of devices that are associated with occupants, the facility may in some embodiments select one device per occupant for which analysis of signals will be performed to determine utilization information. The facility may attempt to identify a "primary" device for an occupant from the cluster of devices associated with the occupant. A primary device may be one that is most often used by the occupant out of the cluster of devices, such as one that is most often carried by the occupant. Signals emitted from a primary device of an occupant may be more indicative of movements of an occupant, and which spaces are occupied by an occupant, than other devices that are not used as often by the occupant.

A utilization analysis facility may select a single device from a cluster of multiple devices associated with an occupant in any suitable manner, as embodiments are not limited in this respect. In some embodiments, signals emitted by devices of a cluster (or information regarding signals, such as the information discussed above in connection with block 304 of FIG. 3) may be analyzed and a single device may be selected based on that analysis. For example, the signals may be analyzed to determine behaviors of the devices in the cluster as exhibited through the signals emitted by the devices. For example, a device that is most active over a period of time out of the devices in the cluster may be selected. As another example, the signals may be analyzed to determine a type of each device and a primary device may be selected based on type. For example, devices of certain types (e.g., mobile phones) that are more likely to be used or carried more often by individuals may be preferred over other types of devices (e.g., laptop computers). As discussed above in connection with FIG. 4, a type of device may be determined from content of signals emitted by a device, such as from the protocol(s) used in the signals or a hardware identifier embedded in the signals.

In embodiments that select a most active device from a cluster of devices, activity may be analyzed according to any suitable criteria. In some embodiments, activity may be evaluated based on a number of signals emitted. In some such embodiments, a most active device over a period of time may be a device that emits the largest number of signals over the period of time as indicated by the number of signals in a set of information regarding signals detected in a space. In other embodiments, activity may be evaluated based on times the device is active, based on an amount of time that a device was continuously or occasionally transmitting signals. In such a case, the device that transmitted signals for the longest portion of a period of time may be the most active device. In still other embodiments, activity may be evaluated based on movements of the device as indicated by estimated locations from which signals were emitted. In such a case, a device that moved to the most number of different locations over a period of time may be the most active device. In other embodiments, some combination of these factors or other factors may be used to determine activity of a device from an analysis of signals emitted by that device and to identify a most active device of a cluster of devices associated with an occupant.

It should be appreciated, therefore, that embodiments are not limited to implementing any specific technique for identifying a single device of a cluster of devices associated with an occupant to identify as a primary device of the occupant, to be used in further analysis of utilization. FIG. 6 illustrates one technique for identifying a primary device of a cluster that may be implemented by a utilization analysis facility in some embodiments.

Prior to the start of process 600 of FIG. 6, information regarding a set of signals emitted by devices and detected in a space over a period of time (e.g., the information discussed above in block 304 of FIG. 3) is received by a utilization analysis facility for analysis. In addition, one or more clusters of multiple devices may have been identified with each of the clusters associated with an occupant. Such clusters may have been identified in any suitable manner, including from an analysis of the information regarding electronic communication signals emitted from a set of devices over time to determine a relationship between devices and occupants of a space. Such clusters may have been identified, for example, using the process 500 of FIG. 5. The process 600 may be used to identify, from within each cluster of multiple devices, a primary device of the occupant associated with that cluster. Accordingly, while the process 600 is described as being performed once for a single cluster, it should be appreciated that the process may be performed multiple times for multiple clusters in some embodiments.

The process 600 begins in block 602, in which the utilization analysis facility analyzed information regarding the signals emitted by the devices of the cluster to attempt to determine a type of each device in the cluster. The signals may be analyzed to determine the type of device in any suitable manner, including using techniques described above in connection with block 402 of FIG. 4. Device type information may be used to determine primary device of an occupant because, in some embodiments, devices of certain types may be preferably identified as primary devices over devices of other types, as discussed above.

In block 604, the utilization analysis facility reviews the signals emitted by each of the devices in the cluster to identify a frequency of movement of each device. The frequency of movement may be identified by a number of locations (within a period of time) from which each device emitted a signal, as indicated by the information regarding detected signals.

In block 606, lengths of movements over a period of time (e.g., over the course of a period of time, such as a day) within a space may be evaluated by the utilization analysis facility. The facility may identify lengths of movements in any suitable manner, such as by determining a sequence of locations within that period of time from which the device emitted signals, determining the distances between each of the locations in the sequence, and summing the distances to determine an amount of distance traveled by the occupant during the period of time as indicated by the detected signals.

The movement information of block 604 and 606 may indicate which device is more likely to be an occupant's primary device, as a device that is brought to more locations (as in block 604) or carried over longer distances (as in block 606) may be the device that an occupant uses as his/her main device and the one more likely for the occupant to carry with him or her as the occupant goes about the day.

In block 608, based on the device type information as well as the movement information collected in block 604, 606, the utilization analysis facility identifies the primary device of the occupant of the cluster. The facility may make the determination based on these factors in any suitable manner, as embodiments are not limited in this respect. In some embodiments, the facility may first rely on whether it was able to identify a specific device type in block 604. As discussed above in connection with FIG. 4, some companies may manufacture devices of multiple different types and, as such, a hardware identifier that identifies one of those manufacturers may not identify a device of a specific type. If, however, the facility is able to determine from the analysis of block 604 that one or more of the devices in the cluster are of a certain type, the facility may review a ranking of devices types to determine any of those determined device types match types that have been pre-identified (e.g., by a user) as device types that should always be used as primary devices. For example, a facility may be configured to always select a mobile phone as a primary device when a mobile phone can be identified from among the cluster of devices for an occupant and the facility may therefore determine whether the device types indicate any mobile phones in the cluster. In these embodiments, in the event that device type cannot be used to determine a primary device, the movement information determined in blocks 604, 606 may be used to select a primary device. For example, the number of locations and lengths of movements for each device may be compared in any suitable manner (in some embodiments, using suitable weighting factors for each variable) to determine a most mobile device among the devices in the cluster. The most mobile device may then be selected as the primary device of the cluster. Once the primary device is identified in block 608, the process 600 ends.

As a result of performing the process 600 on one or more clusters of devices, particularly in combination with techniques for filtering/blacklisting devices as discussed above in connection with FIG. 4, a limited set of devices can be identified that are uniquely associated with occupants among the devices in the set. In addition, as the devices have been selected to be the "primary" devices for the occupants in the case that multiple devices were identified for an occupant, the devices of the limited set may be the ones that the facility has identified as likely to move with occupants and therefore the devices for which signals emitted by the devices will be most indicative of the presence of the occupants in a particular space. The devices of the limited set may therefore be those for which signals emitted by the devices are most indicative of utilization of a space. Accordingly, a utilization analysis facility may use these devices and information regarding signals emitted by these devices over a time period to determine utilization of a space, including sub-spaces of that space.

Information regarding signals emitted by devices of a limited set may be analyzed in any suitable manner to determine utilization of a space, as embodiments are not limited in this respect. FIG. 7 illustrates an example of a process 700 that may be used in some embodiments to determine utilization of a space based on analysis of signals.

Prior to the start of the process 700 of FIG. 7, information regarding a set of signals emitted by devices and detected in a space over a period of time (e.g., the information discussed above in block 304 of FIG. 3) is received by a utilization analysis facility for analysis. In addition, a set of occupants and a set of devices uniquely associated (among the set of devices) with those occupants may have been identified. Through the process 700, a utilization analysis facility may determine utilization of the space through analyzing the information regarding the signals emitted by the devices of the set.

The process 700 may be used to determine utilization of a specific space during a specific period of time through determining utilization of that space in successive time intervals within the specific period of time. The time intervals may be any suitable division of time within the period of time of interest. For example, the time period of interest may be a day and the time intervals may be 15 or 30 minutes. In some embodiments, utilization information may be determined for multiple different spaces and sub-spaces during the time period of interest. The process 700 may be used in such embodiments to determine the utilization of a particular space or a particular sub-space during the time period of interest. As discussed in more detail below, a utilization analysis facility may aggregate information on utilization of a specific space during a specific period of time with utilization information for other spaces to produce information about utilization for larger spaces. As illustrated in FIG. 7, the process 700 includes iteratively determining utilization of a space over the course of successive time intervals, which are fractions of a time period of interest. Accordingly, at a start of the process 700, a utilization analysis facility may begin analyzing utilization of the space during a first time interval of the time period of interest, then proceed to the next time interval, and so on.

The process 700 begins in block 702, in which a utilization analysis facility analyzes electronic communication signals emitted by devices of the set before the time interval. More specifically, the facility may analyze signals that were emitted before the time interval and for which localization information for the signals indicates that the signals were emitted by devices that were located within the space.

The facility may identify signals emitted by devices inside the space in any suitable manner. For example, the facility may make the identification through comparing coordinates of estimated location from which each of the signals was emitted (which may have been determined, as discussed above, using known localization techniques, including known triangulation techniques) to coordinates for the space. As discussed above in connection with FIG. 3, as part of an initialization for the utilization analysis facility, coordinates for the space may be defined, which may include determining a polygon that represents the space in the coordinate system. As also discussed above, any suitable coordinate system may be used, including a geographic coordinate system, as embodiments are not limited in this respect. To determine whether a signal was emitted by a device in the space, the utilization analysis facility may determine whether the coordinates of the estimated location of the device that emitted the signal fall within the polygon for the space. If so, and if that signal was emitted before the time period of interest, the signal may be reviewed by the facility in block 702.

It should be appreciated that in reviewing signals emitted before a time interval, the facility may review signals emitted in any suitable amount of time before the time interval. For example, the facility may review signals that were emitted within some threshold time before the time interval, such as within a certain number of minutes, hours, or other length of time before the time interval. Such a threshold may be determined in any suitable manner, including as a fraction of the time interval or an amount of time equivalent to the time interval, or as an absolute amount of time input by a user. As another example, the facility may review signals that were emitted in the time interval prior to the time interval currently under evaluation by the facility. It should be appreciated that, in some cases, for a first time interval of a period of time, information about signals emitted before the time interval or in a preceding time interval may not be available. The signals may be analyzed in any suitable manner in block 702. For example, in some embodiments the utilization analysis facility may simply note the presence of the signals, and/or the presence of the device, in the space before the time period of interest. In other embodiments, the facility may analyze a length of time the signals were emitted or other characteristics of the signals.

In block 704, the utilization analysis facility reviews signals that are emitted by devices (of the set of devices) inside the space and during the time interval, and in block 706, the facility reviews signals emitted by devices inside the space and after the time interval. In blocks 704 and 706, the facility may determine the presence of devices inside the space in the same manner as discussed above in block 702, and may carry out a similar review of the signals. In block 704, for example, the facility may identify a proportion of the time period of interest for which signals from each device were detected. For example, if the time interval is 30 minutes, the facility may review signals emitted by a device inside the space and determine that the device transmitted signals from within the space for 10 of those 30 minutes, or for one-third of the time interval. In block 706, the analysis of signals emitted after the time interval may be those signals emitted during the amounts of time discussed above in connection with block 702.

In block 708, the utilization analysis facility may determine utilization information for the space during the time interval based on the signals reviewed in blocks 702-706. The facility may use each of the signals discussed above - both signals detected during the time interval as well as signals detected before and after the time interval - due to inherent uncertainties of electronic communication signals as they relate to utilization. As mentioned above, it is occasionally the case that a device is present in a space before, during, and after a time interval, but does not emit any signals during that time interval. If the facility only analyzed signals emitted during the time interval, the facility would erroneously conclude that the device was not present in the space during the time interval, and erroneously conclude that the occupant who operates the device was not present in the space during the time interval. Though, that device may transmit signals both before and after the time interval. Using techniques described herein, the facility may conclude that because the device emitted signals before and after the time interval, the device was likely also present in the space during the time interval.

Through analyzing signals emitted from a space both before, during, and after a time interval, the utilization analysis facility may classify the relationship between each of the devices and the space for which utilization is to be determined into one of eight categories that relate to emission of signals in the space with respect to the time interval: (1) Devices that emitted signals detected in the space before, within, and after the time interval;

(2) Devices that emitted signals detected in the space before and after the time interval, but not during the time interval;

(3) Devices that emitted signals detected in the space only during the time interval;

(4) Devices that emitted signals detected in the space only before and during the time interval;

(5) Devices that emitted signals detected in the space only during and after the time interval;

(6) Devices that emitted signals detected in the space only before the time interval;

(7) Devices that emitted signals detected in the space only after the time interval; and

(8) Devices that did not emit signals detected in the space before, during, or after the time interval. The utilization analysis facility may determine utilization of the space during the time interval in block 708 using these eight categories. Some of the categories may be indicative of utilization of the space during a time interval and other categories may not be indicative of utilization of the space during the time interval. Accordingly, when a device is categorized into one of the categories indicative of utilization, the facility may infer from that categorization that an occupant who operates that device was present in the space for at least a portion of the time interval and may increase utilization of the space during the time interval accordingly.

In some embodiments, the facility may determine utilization in a binary basis, in that the facility may determine whether an occupant was present in the space during the time interval or not and set utilization based on that determination. In such embodiments, the facility may review a categorization of devices with respect to the space to determine which devices emitted signals indicative of the devices being present in the space during the time interval. For example, using the categories above, categories (1) - (5) may be identified by the facility as being indicative of presence in the space during at least a portion of the time interval. The facility may count the number of the devices that are categorized into categories (1) - (5) and, because each device is uniquely associated with one occupant, assume that the count of devices is equal to the count of occupants.

In other embodiments, the facility may determine utilization in a manner that accounts for presence of an occupant in a space during only a portion of a time interval. For example, the utilization analysis facility may attempt to determine from an analysis of the signals whether a device was likely present in the space for an entirety of a time interval or a portion of the time interval. The facility may then determine utilization based at least in part on the presence in the space during the time interval as well as the amount of time the device spent in the space during the time interval, as inferred from the signals emitted by the device and detected in the space. Such an utilization determination may be termed a duration-weighted utilization.

In embodiments that calculate a duration-weighted utilization for a space during a time interval, the facility may determine the utilization from the categorization of devices discussed above. Each of the categories may contribute a different amount to a calculation of utilization. In the calculation, a value of 1 may indicate that a device (and its occupant) was inferred to have been present in the space for an entirety of the time interval, while a value that is a fraction of 1 may indicate that the device was inferred to have been present in the space for that fraction of the time interval. Accordingly, for each device, the facility may calculate a value for utilization of the space by that device that is in proportion to an amount of the time interval that the device is inferred to have been present in the space.

Specifically, as should be appreciated from the discussion above, categories (1) and (2) above may be indicative of the device having spent an entirety of the time interval in the space. When a device is categorized into (1) or (2), the facility may add a full value 1 to a count of occupants in the space during the time interval, because the device was inferred to have been present for an entirety of the time interval and therefore utilized the space for the entirety of the time interval. Categories (6) - (8) may be indicative of the device not having spent any of the time interval in the space, and the facility may accordingly, for each device categorized into these categories, add a value of 0 (or take no action) to the number of devices present in the space. Categories (3) - (5) indicate that a device was present in the space for at least a portion of the time interval: the device may have been in the space before the time interval and left during the time interval, entered and left the space during the time interval, or entered the space during the time interval and left after. In each case, the facility may analyze signals emitted by devices emitted by devices categorized into these categories to infer an amount of time that the signals indicate that the device was present in the space during the time interval. For example, from the signals analyzed in blocks 702 - 706, the facility may determine a time of a first signal emitted by a device and a time of a last signal emitted by the device to determine a time range of the signals and then compare this time range to the time interval to determine a portion of the time interval overlapped by the time range. In some cases, multiple time ranges may be considered, such as where the signals indicate that a device enters the space, leaves, and then enters again during the time interval. A value for utilization of the space by that device may then be a fraction that is proportional to the amount of the time interval overlapped by the time range. For example, if the analysis indicates that the device was transmitting signals in the space for 20 minutes out of a 30 minute time period, the utilization value for that device may be calculated as 20/30 = 0.67.

The facility may sum each of the utilization values determined for each of the devices (0, 1 , or a fraction) to determine a total count of devices present in the space during the time interval. In a calculation of duration-weighted utilization, this may be a value that is not a whole number. As discussed above, the facility may then infer the count of devices present in the space to be equivalent to a count of the number of occupants in the space, due to the relationship between devices and occupants.

As discussed above, utilization is a ratio of a number of persons present in a space to a capacity of a space. Accordingly, in block 708 facility may determine from information regarding the space a capacity of the space and calculate utilization as a ratio of the number of occupants (i.e., inferred persons) present in the space, as determined using the techniques discussed above, to the capacity of the space. Once the utilization is determined for the space for the time interval, the utilization may be stored in a data store for future reporting.

In block 710, the utilization analysis facility determines whether any more time intervals of the time period of interest are to be analyzed. If so, the facility loops back to perform the analysis of blocks 702 - 708 for the next time interval. If not, the facility continues to block 712 to determine utilization of the space during the time period of interest. The utilization of the space over an entirety of the time period of interest may be a function of the utilization of the space during the time intervals, which are portions of the time period of interest. For example, the utilization of the space over the entirety of the time period of interest may be calculated as an average of the utilization during the time intervals.

As discussed above in connection with FIG. 3, as part of determining utilization, a utilization analysis facility may determine a utilization of a space over a particular time period, and may additionally determine a number of occupants in the space during the time period. The facility may also determine any relevant statistics regarding the utilization or number of occupants in the space during a time interval, which the facility may determine through analyzing utilization over a period of time longer than the time interval.

Once the utilization of the space over the course of the time period of interest is determined in block 712, the facility may store utilization in a data store for future reporting to a user, and the process 700 ends.

In the discussion of the analysis of signals in blocks 702 - 706 above, the facility was described as analyzing the locations identified for each of the signals emitted by devices during a time period of interest to determine whether a device was present or not within a space. In some embodiments, the individual locations for each of the signals may be analyzed by the facility. Such an embodiment may be used in scenarios in which multiple different sub-spaces of a space are to be analyzed and the facility is to determine a portion of each time interval that devices spend in each space. Such a determination can enable a fine-grain analysis of utilization of each sub-space. However, those skilled in the art will appreciate that such a determination may be computationally expensive. Further, such a determination may be inherently imperfect when some localization techniques are used to determine the locations of devices that emitted signals, due to the imprecision of such localization techniques. In other embodiments, accordingly, the utilization facility may simplify a determination of utilization by analyzing the various locations of each device during each time interval as indicated by the signals emitted by that device during that time interval. For each device, the facility may determine a single location to use as an approximation of the device's location during the time interval, and may use that location as the sole location of a device during the time interval.

FIG. 8 illustrates a process 800 that may be used in some embodiments for determining a single location of a device during a time period based on signals emitted by that device during the time period and estimated locations of the device at the times those signals were emitted.

The process 800 begins in block 802, in which a utilization analysis facility analyzes information regarding signals emitted by a device over time (e.g., the information described above in connection with block 304 of FIG. 3) to identify the signals emitted by the device during the time period of interest. In addition, the facility determines from that information the emission time for each signal and the estimated location of the device at the time each signal was emitted. In block 804, the facility processes those times and those locations to identify an amount of time that the device spends in each location. Based on that analysis, the facility may determine an average location of the device during the time period, weighted by the duration the device spends at each location indicated for the signals. The average location may be an average coordinate (e.g., average (x,y) coordinate, including geographic coordinate) based on the coordinates indicating for each of the signals. In block 806, that duration-weighted location may be output by the utilization analysis facility as the single location of the device during the time period, and the process 800 ends. Following the process 800, the duration-weighted location of the device may be used in any suitable manner, such as in identifying a position of a device during a time interval so as to determine utilization of spaces during the time interval, as discussed above in FIG. 7.

In the discussion of FIGs. 7-8 above, utilization of a space was determined based on an evaluation of information regarding electronic communication signals emitted by devices of a limited set, where those devices had been determined to be "primary" devices operated by each occupant of the space. In the technique as described, devices operated by an occupant other than that occupant's primary device were not considered and signals emitted by those other devices were not considered. It should be appreciated, however, that embodiments are not limited to evaluating only signals emitted by a limited set of primary devices. In some embodiments, a utilization analysis facility may evaluate signals emitted by two or more devices operated by one occupant in determining utilization of a space.

For example, while an occupant may primarily carry his/her primary device when moving, there may be times when the occupant may leave the primary device in one location and not carry it to another and may instead carry another device. Additionally, there may be times when a primary device is not indicative of a position or movement of an occupant, as the primary device may not be emitting signals or may be powered off. Under such circumstances, if the utilization analysis facility were to only consider signals emitted by the primary device, the facility could misreport utilization by erroneously reporting that the occupant stayed in one position when the occupant left a primary device behind or erroneously reporting that the occupant was not in a space when the primary device was not emitting signals while the occupant was in that space.

To limit the effects of such errors, in some embodiments the utilization analysis facility may analyze movements of the primary device and determine whether, during a time period, to track presence or movements of an occupant using a device other than the primary device. In such embodiments, the facility may have previously identified a set of two or more devices that are operated by the occupant (such as using the technique of FIG. 5 above) and the facility may select a device from that set to use in tracking presence or movements of the occupant. More specifically, in some such embodiments the utilization analysis facility may evaluate signals emitted by the primary device operated by the occupant to determine whether, for a time period, the signals indicate that the primary device did not move or did not emit signals. If the facility determines that the primary device did not move or did not emit signals for the time period, the facility may determine whether any of the other devices operated by that occupant emitted signals during that time period. If one of the other devices operated by that occupant emitted signals during that time period and the signals indicate that the occupant moved during the time period, then the facility may select that one device to monitor for the time period. If more than one of the other devices operated by that occupant emitted signals during that time period, then the facility may select between those devices to choose one to monitor for the time period, such as by choosing the device that was the most active during the time period. Once the one device is selected, the facility may for that time period track the presence and movements of the occupant using that one device and, accordingly, track utilization of a space based on signals emitted by that one device. Following the time period, the facility may return to tracking the occupant using the primary device and determining utilization of a space based on signals emitted by the primary device.

As should be appreciated from the foregoing, utilization is a function of capacity of a space. Accordingly, a determination of utilization of a space can be affected by how that capacity is measured. Different spaces may have sub-spaces of different types, each of which may have a capacity. For example, as discussed above in connection with FIG. 2, spaces may be categorized as workstation spaces, support spaces, executive spaces, and other types of spaces. In some cases, an organization may be interested in determining utilization of its space, but may not be interested in accounting for some parts of that space in the determination of utilization. For example, when determining utilization of an office building as a whole, a business may decide not to evaluate utilization of its executive offices, as the business may have determined that the executive offices are not going to be changed in any way as a result of the analysis. Similarly, a business may have determined that its lobby or reception area are not going to be changed and may determine that utilization information for these areas is not needed. In such cases, when spaces such as these are excluded from the analysis, the capacity of these spaces is also removed from the final calculation of utilization. Removing the capacity of these spaces may increase or decrease a reported utilization value. For example, if a reception area was designed with a high capacity but is often unused, that high capacity combined with low presence of occupants in the space would lower a reported utilization. By excluding the lobby, a utilization value for an office that includes that reception area may increase, and this utilization value may be more tailored and useful to the business that operates that office.

In some embodiments, a utilization analysis facility may accept user input specifying types of spaces to be considered in determining utilization, and determine utilization based in part on that input. FIG. 9 illustrates an example of a process that may be performed in some embodiments for determining utilization based in part on user input regarding types of spaces to be considered.

The process 900 of FIG. 9 begins in block 902, in which an initial configuration of a utilization analysis tool is performed by classifying each sub-space identified in a floor plan of a space. The sub-spaces and the space may be of any suitable type, as embodiments are not limited in this respect. In some scenarios, the space may be an office (e.g., the office 200 of FIG. 2) and the sub-spaces may be rooms within that office (e.g., the rooms 202-212 of FIG. 2). The user may provide the input in any suitable manner, such as by viewing the floor plan in a graphical user interface and classifying each of the sub-spaces according to the illustrative classifications discussed above in connection with FIG. 2. The classifications may be stored in a data store along with other information about the floor plan and about the sub-spaces. A data store of information about the sub-spaces may additionally include coordinates defining each sub-space within the floor plan and capacity information for each sub-space, both of which may also have been entered by the user via the graphical user interface.

In block 904, following the configuration of block 902, a utilization analysis facility may receive input from the user requesting that utilization of the space be determined, and specifying one or more classifications of sub-spaces that should be included in the utilization determination. As a result of the input of block 904, in block 906 the facility may begin analyzing utilization of sub-spaces of the space during a time period of interest. The sub-spaces that are analyzed in block 906 may be those that are classified as being of types that correspond to the types specified by the user input of block 904. The facility may determine the utilization of the spaces in any suitable manner, including using the process 700 of FIG. 7 discussed above.

In addition to determining the utilization of the sub-spaces in block 906, in block 908 the facility determines utilization of the space as a whole during the time period of interest. That utilization value will be affected by the user input of block 904, in that the capacity of the space as a whole that is used in determining the utilization is determined based on the sub-spaces that meet the classifications specified by the user. Specifically, the capacity may be determined as a sum of the capacities of the sub-spaces that meet the classifications specified by the user. The number of occupants in the space as a whole during a time period of interest may be determined in any suitable manner. In some embodiments, the utilization of the space as a whole may be determined as a function of the utilizations of the sub-spaces. As another example, the utilization of the space as a whole may be calculated anew, such as using the duration-weighting techniques described above in connection with FIG. 7 and, in some cases, the average location techniques described above in connection with FIG. 8. For example, the space as a whole may be considered during the time period of interest without regard to the division of the space into sub-spaces, and the process 700 of FIG. 7 may be used to determine a number of occupants within the space during the time period of interest. In some such embodiments, the occupants may only be counted as being present in the space when signals indicate that their devices are within the portions of the space corresponding to the sub-spaces of the classifications specified by the user. That count of occupants present in the space during the time period may then be compared to the capacity of the space, which the facility may have determined as a sum of the capacities of the sub-spaces. The facility may then determine a utilization value from the count of occupants and the capacity. The facility may then output the utilization values for the space and for the sub-spaces in block 908, which may include outputting the values for display via a graphical user interface or outputting the values for storing in any suitable data store. Once the values are output in block 908, the process 900 ends. Using the illustrative processes of FIGs. 4-9, information regarding a set of electronic communication signals detected in a space over a period of time may be used to infer utilization of that space. As should be appreciated from the foregoing discussion of those illustrative processes, a utilization analysis facility may carry out machine learning and computerized analysis techniques to filter the set of signals emitted in the space down to a relatively small number of signals, by excluding from consideration signals emitted by various devices, which can aid in the determination.

As mentioned above in connection with FIG. 3, before these filtering and analysis processes are carried out, in some embodiments the utilization analysis facility may first analyze the signals detected in the space to filter the data, to remove from consideration data that is erroneous or irrelevant. Such a filtering process may be carried out in any suitable manner, as embodiments are not limited in this respect.

FIG. 10 illustrates one example of a process that a utilization analysis facility may carry out to perform such filtering. Prior to the start of process 1000 of FIG. 10, information on a set of signals emitted by devices over a period of time and detected in a space is collected. This information may be of a same type of information as discussed in connection with block 304 of FIG. 3. The information may indicate one or more signals emitted by each device during the period of time, and may indicate for each device a received signal strength (RSS) for that signal, an identifier for the device that emitted the signal, and an estimated location of the device that emitted the signal, among other information. In addition, prior to the start of the process 1000, a user may have specified a space for which utilization is to be determined, where the space corresponds to the space in which the signals were detected.

The process 1000 begins in block 1002, in which a utilization analysis facility analyzes the information about the signals to identify signals for which the information indicates an estimated location outside of the space. The facility may determine whether an estimated location listed for a signal is outside of the space in any suitable manner, including using the technique discussed above for comparing coordinates of an estimated location to coordinates of a space. The facility may respond to a determination that a signal was emitted by a device outside the space by filtering that signal from further analysis, including by editing the information regarding the signals to delete the information for that signal. The facility may repeat the determination for each signal in the set and filter out all signals for which the estimated location is outside the space.

In block 1004, the facility may additionally attempt to identify devices located outside the space through reviewing a RSS for each signal. Due to the imprecision of some localization techniques, it is possible that a signal that was emitted by a device outside a space may be erroneously indicated to be within the space. However, a RSS of such a signal may be low, due to the signal traveling the distance from outside the space to the inside of the space (and through any intervening materials, such as walls, floors/ceilings, etc.). Accordingly, without consideration of the location indicated for a signal, if the facility determines that information for a signal indicates an RSS below a threshold, the facility may filter that signal from further analysis, such as by deleting records regarding the signal.

In block 1006, the facility may review the signals emitted by each device indicated by the information for signs of erroneous data in the signals, by comparing information about the signals emitted by a device to information about other signals emitted by that device. For example, the facility may determine whether the information for a signal indicates an erroneous location for the signal, which the facility may determine by evaluating in context with other locations for one or more other signals emitted by that device within a threshold period of time. For example, the facility may compare signals emitted by a device within a range of time to determine whether any one or more of the signals indicate a movement of the device that is greater than a threshold distance. Due to imprecision of some localization techniques, for example, a location for a signal may be erroneously reported and the erroneous location may be a long distance away from other signals emitted close in time that may have had correct locations determined. A large variation in distance may be at best improbable and at worst impossible, and the signal having the location that is more than the threshold distance from the other signal(s) may be filtered by the facility. In some embodiments, the threshold may be expressed as a distance, while in other embodiments the threshold may be expressed as a speed (i.e., distance over a period of time). In embodiments that consider a speed, both a distance between locations and a difference in times for two signals may be used to determine whether a signal is erroneous for indicating that a device would have had to have moved more than a threshold speed for the locations and/or times to be correct.

In some embodiments, in addition to analyzing individual signals in block 1006 to determine whether the signals are erroneous, the utilization analysis facility may track a number of erroneous signals identified for each device. In a case that the facility identifies more than a threshold number of erroneous signals for a device, the facility may flag that device as an untrustworthy source of data and may exclude all signals emitted by that device from further analysis.

In block 1008, the utilization analysis facility may additionally evaluate signals emitted by each device over time and apply a smoothing algorithm to the signals for each device. The smoothing algorithm may be used to merge signals and condense the data for each device, so as to simplify the data for each device and ease analysis. Any suitable smoothing algorithm may be used in embodiments that apply a smoothing algorithm, as embodiments are not limited in this respect. In some embodiments, a Savitzky-Golay or a Moving Average smoothing algorithm may be performed over a series of iterations to analyze the locations indicated by the signals emitted by a device and edit the data to smooth a travel path of the device as indicated by the locations of signals over time, by adjusting the individual locations identified for the emitted signals. More specifically, over the course of several iterations, signals that are similar to one another may be merged to create a merged set of information about the signals (e.g., an average may be calculated for the signals) or may be filtered such that only one of the similar signals is used in further analysis.

Once the smoothing algorithm is performed to smooth the travel path of each device in block 1008, the process 1000 ends.

Using the techniques of FIGs. 3-10, utilization of specific spaces may be determined. That information may be used in any suitable manner by organizations, including in identifying how well their space is meeting their needs and whether adjustments to the space might be needed or advisable. In some cases, to aid organizations in determining whether their space is meeting their needs or whether adjustments could be made, in addition to tracking utilization of spaces by occupants during one period of time, the utilization analysis facility may classify the occupants according to how their behaviors affect utilization of the space. For example, the facility may analyze presence in the space by each individual occupant over a longer period of time. Based on that presence, the facility may classify the occupant into one of a set of behavior- based classifications.

Any suitable classifications may be used, as embodiments are not limited in this respect. In some embodiments, mobility classifications may be used that reflect amounts of time the occupant spends in the space or spends outside the space, or reflect behaviors exhibited by the occupant inside the space. For example, classifications used in some embodiments that relate to a commercial office space may include a classification that relates to an occupant who spends nearly all of his/her time in the office at a single location (which can be inferred in this context to be the occupant's desk); a classification for an occupant who is typically in the office during working hours, but often moves within the office during the day; and a classification for an occupant who usually splits time roughly evenly between working in the office and working outside the office.

Each of these exemplary categories may be associated with specific characteristics relating to behaviors that may be exhibited by occupants. For example, each category may be associated with a threshold average number of days per week spent in a space, threshold average number of hours per day spent in a space, and/or threshold average number of unique locations within the space per day as exhibited by signals. The "deskbound" category, for example, may be defined as spending more than 3.5 days per week in a space on average and having no more than one unique location within the space on an average day. Similarly, the "internally mobile" category may be defined as spending more than 3.5 days per week in a space on average, but having more than one unique location within the space on an average day. Any suitable set of thresholds and other conditions may be used, as embodiments are not limited in this respect.

FIG. 11 illustrates an example of a process 1100 that may be used in some embodiments for categorizing occupants into a set of categories based on behaviors exhibited by those occupants. Prior to the start of the process 1100 of FIG. 11, information regarding a set of signals emitted by devices and detected in a space over a period of time (e.g., the information discussed above in block 304 of FIG. 3) is received by a utilization analysis facility for analysis. In addition, a set of occupants and a set of devices uniquely associated (among the set of devices) with those occupants may have been identified. Through the process 1100, a utilization analysis facility may determine behaviors of an occupant through analyzing the information regarding the signals emitted by the device in the set that is associated with that occupant. The process 1100 can be repeated for each device in the set, so as to classify each of the occupants.

The process 1100 begins in block 1102, in which the utilization analysis facility analyzes the information regarding the signals to determine an amount of time an occupant spends within a space and to determine movements of the occupant within the space. The facility may perform the analysis over a larger period of time to determine average amounts of time and average movements for the occupant within smaller periods of time, such as by analyzing signals detected over the course of one or more months so as to determine average behaviors for a day and/or a week. The facility may carry out the analysis by determining from the times and estimated locations for each of the signals emitted by the occupant's device to determine any suitable measure of behaviors of the occupant within a space. For example, the facility may determine an average number of distinct locations visited by an occupant during a day, an average number of hours spent in the space by the occupant during a day, and an average number of days the occupant spends at least a part of a day in the space per week.

In block 1104, the facility may also analyze the information regarding signals collected over the same period of time from block 1102 to identify the absence of signals emitted by the device and thereby determine behaviors of the occupant. More specifically, the facility may analyze times that the occupant's device did not emit signals detected in the space, as indicated by the information regarding the detected signals. Absence of signals may be indicative of the occupant not being present in the space. The facility may analyze the lack of signals to determine any suitable behaviors relative to a period of time, including an average number of hours not spent in the space by the occupant during a day and an average number of days the occupant does not spend in the office.

In block 1106, the utilization analysis facility uses the results of the analysis of blocks 1102 and 1104 to classify the occupant based on the inferred behaviors. The facility may perform the classification in any suitable manner, including according to one or more thresholds defined for each classification. For example, a classification may be associated with a threshold amount of time spent in the space over a period of time (or two or more thresholds each associated with different periods of time) and/or a threshold amount of activity within the space, such as a threshold number of locations visited in the space. The thresholds may, in some embodiments, be associated with average values for periods of time.

As another example of a way in which the utilization analysis facility may perform the classification of block 1106, in some embodiments a machine learning algorithm may be used to identify a classification for an occupant based on signals emitted by that occupant's device. For example, using a set of devices that have been manually labeled with particular classification of occupants that operate those devices and a set of signals exhibited by those labeled devices, a classifier may be trained for each classification with characteristics of signals emitted by devices/occupants of that classification. Such characteristics may include any suitable information that may be inferred from the signals, such as timing or location characteristics that may relate to times at which the devices emit signals, locations at which the devices emit signals, movements of the devices as exhibited by the signals, protocols used by the devices, etc. Once the classifiers are trained using the labeled data, the machine learning algorithm may be used to determine a classification of an occupant based on signals emitted by that occupant's device. More specifically, the utilization analysis facility may review the signals emitted by a device using the machine learning algorithm to determine which classifier most closely matches that device, and thereby that occupant. Those skilled in the art will appreciate how to implement a machine learning algorithm to carry out such an analysis.

Once the occupant is classified in block 1106, the process 1100 ends. The classification may be stored by the utilization analysis facility in a data store and used in any suitable manner. For example, once each of the occupants have been classified using the process 1100, the results may be output to a user. The user may be able to use the information to identify behaviors of occupants, such as behaviors of employees of a business, to make determinations about how the space is satisfying needs or how adjustments could be made to the space. For example, by determining through the classification how many employees of a business work remotely for part of an average week and how many employees come to work each day, the business may be able to determine whether an increase or decrease in a number of offices would be advisable. The process 1100 of FIG. 11 was described as determining a single classification for an occupant based on behaviors of that occupant and how the behavior affects utilization of a space. It should be appreciated that behaviors of an occupant may change over time and, therefore, a classification of an occupant may also change over time. Accordingly, in some embodiments, the classification of an occupant based on behavior may be performed based on behavior exhibited during a time period and a classification of an occupant may vary between time periods.

Techniques have been described for inferring utilization from an analysis of electronic communication signals detected in a space. As should be appreciated from the foregoing discussion, however, embodiments are not limited to making utilization determinations based only on electronic communications signals. In some embodiments, in addition to or as an alternative to analyzing electronic communication signals, a utilization analysis facility may analyze other information indicative of the presence of occupants in a space.

Any of various kinds of information may be analyzed in embodiments that are indicative of the presence of occupants in a space. The information indicative of the presence of occupants in a space may be any suitable information in any suitable format, as embodiments are not limited in this respect. For example, such information indicative of the presence of an occupant in a space may include information indicating usage of a network-connected device in the space, such as usage of a hardwired device such as a computer or Voice-over-Internet- Protocol (VoIP) phone that is connected to a wireline network port in the space. Information indicative of the presence of an occupant may also include physical occupancy sensors, which may include sensors installed in desks or in seats to detect the presence of an individual. Information indicative of the presence of an occupant may also include cameras and image processing engines that are configured to detect the movements of occupants, such as cameras installed above or near entrances to spaces. As discussed above, presence may also be indicated by requests for access for a space, such as requests for physical access or requests for electronic access.

Embodiments may use information indicative of presence of individuals in any suitable manner. In some cases, such information may indicate a sequence of events relating to an occupant in a space. As a specific example, information about usage of a network-connected device may indicate times that multiple communications were sent over a wireline network port or a range of times that communications were sent over a network port. As another specific example, information produced by a system including a camera may include times indicating entrances of occupants into a space. In cases where the information indicates the occurrence of multiple events over time, and each event identifies the presence of an occupant in a space and a time of the event, a utilization analysis facility may identify a first event for each occupant within a time period and identify the time of the first event as a time that that occupant arrived in the space. As discussed in more detail below, the facility may additionally predict a length of time that each of the occupants will spend in the space and thereby identify departure times from the space for each of the occupants. Using the inferred arrival times and predicted departure times, the facility may track utilization of the space over time.

Specific examples of information that may be indicative of the presence of occupants in a space include requests for access for a space made by occupants. It should be appreciated, however, that embodiments that operate with information indicative of the presence of occupants in a space are not limited to operating with requests for access for a space.

FIG. 12 illustrates an example of a process 1200 that may be used in some embodiments to determine utilization information for a space based on information indicative of the presence of occupants in a space. The process 1200 begins in block 1202, in which a utilization analysis facility receives information indicative of the presence of individuals in a space. In the example of FIG. 12, such information may be in the form requests for access. A request for access may have been made by an individual in any of various ways, including through a presentation of credentials to demonstrate permission for access to be granted. For example, the requests for access could be requests for physical access to the space, in which case the presentation of credentials may include scanning/swiping of a tangible credential (e.g., an employee identification badge that is presented using a bar code, RFID, or other technology), providing a biometric credential (e.g., a fingerprint scan, retina scan, voiceprint scan, etc.). As another example, the requests for access may be requests for electronic access (e.g., to a computer network) made from a device within a space, and the presentation of credentials may include an input of electronic credentials like a username or password, biometric credentials, or any other suitable credential.

In some embodiments, including the embodiment of FIG. 12, the information that indicative of presence may be information from which a utilization analysis facility may infer arrival of an occupant to a space. A request for physical access typically immediately precedes presence in a space. A request for virtual access indicates that an occupant is beginning use of a computing device in the space, which often occurs soon after the occupant arrives in the space.

Accordingly, the receipt of information in block 1202 may include receipt of a set of information indicating a set of arrival events that each indicate arrival of an occupant to a space. The information may be in any suitable format and include any suitable information, as embodiments are not limited in this respect. In some embodiments, the information for each arrival (i.e., for each request of access) may identify the occupant that arrived in the space. The occupant may be identified in any suitable manner, including by a name, identification number (e.g., employee number, credential number), demographic information (e.g., name, age, gender, etc.), occupation information (e.g., department, job title, etc.), a relationship of the occupant to an organization that operates a space (e.g., whether the occupant is an employee of a business, is a visitor/contractor, etc.), or any other suitable information about the individual. In some embodiments, the information for each arrival may include a time of arrival at the space. The information may also identify the space in some embodiments. The space may be identified directly by the information in some embodiments, while in other embodiments the space may be identified indirectly. For example, in embodiments in which requests for physical access are analyzed, a device (e.g., a credential authentication device) may be used to read an individual's credentials, such as an RFID reader, bar code scanner, or other credential-reading device. In some such cases, the information about a request for access may include information on a device that received the request for access. The utilization analysis facility may be able to determine a relationship between devices and spaces, such as by reviewing data indicating a correspondence between the devices and the spaces to which the devices grant access, and thereby determine from an identification of the device the space to which access was requested. As another example of the space being directly or indirectly identified, in the case of a request for electronic access, the information about the request may include an identification of a space from which the space was received or an indication of a location from which the request was received. In cases where the request for electronic access was transmitted wirelessly, the location may be a triangulated location of a device that emitted the wireless signal, as in techniques discussed above. In other cases where the request was transmitted over a wire, the location may be identified by a network port by which the request was received. The utilization analysis facility may consult information about network ports or locations from which signals were emitted to information regarding spaces to determine the space from which the request for electronic access was received.

The information received in block 1202 may have been received from any suitable source in any suitable format. For example, in some embodiments the information may have been output by a security system that reviewed the requests for access and either approved or rejected the requests.

In some embodiments, in block 1202 the utilization analysis facility may clean and filter the data to remove data that is not indicative of an arrival in a space, such as by removing information regarding requests for access that do not correspond to an arrival of an occupant. For example, if a request for physical access was rejected, an occupant may not subsequently enter a space and, accordingly, the utilization analysis facility may filter such a rejected request from further consideration. As another example, credential-scanning devices may occasionally inadvertently scan an occupant's credentials twice in rapid succession, such as by scanning an RFID badge twice during a single presentation of the badge. Accordingly, the facility may review the received data in block 1202 to identify requests for access that occurred within a threshold amount of time of one another and filter the second request from further consideration.

In embodiments that process information indicative of the presence (including arrival) of an occupant in a space, the utilization analysis facility may analyze the information indicating presence/arrival of an occupant to infer a time at which the occupant may depart the space. The inference regarding departure may, in such embodiments, be drawn based on an analysis the information indicating arrival without an analysis of separate information indicative of departure. In such embodiments, the utilization analysis facility may determine utilization of a space from the information indicative of arrival of an occupant in the space and the inferred departure of the occupant from the space.

Accordingly, in block 1204, the facility reviews the requests for access for each occupant and determines, for that occupant, an inferred time that the occupant will depart the space. In some embodiments, the facility may review a set of requests for access for an occupant for a time period (e.g., a day) and select a first request for access for that time period to infer a time of departure by the occupant only from an analysis of that first request. In such a case, using the example of a day, the facility may identify from a first request for access a time at which an occupant arrived in a space first for the day and may infer a time that the occupant will depart the space for the day. For example, an occupant may be required to request access to an overall space, such as a building, and may over the course of the day request access to various sub- spaces of the space, such as rooms of the building. The facility may review the first request for access to the larger space for the day (or the first request in any of the sub-spaces for the day) and then infer a time that the occupant will depart the larger space.

The facility infers a departure time in any suitable manner, as embodiments are not limited in this respect. In some embodiments, the facility may make the inference using one or more distribution of lengths of time an occupant may spend in a space. Such a distribution may be a probability distribution in some embodiments, such that the facility may determine the departure time probabilistically. In embodiments in which the facility may use a distribution, the facility may select at random a value between 0 and 1 and determine a length of time associated with that value in the distribution. The facility may then use that length of time as the length of time that an occupant will spend in a space. The facility may then predict that an occupant will depart a space after that length of time. In some cases, this determination may be made based only on a first request for access and, as a result of the random selection, the predicted departure time may be before some of the requests for access made by the occupant. In such cases, the predicted departure time may be used regardless. In other embodiments, a facility may respond to determining that a predicted departure time is before another request made by that occupant by adjusting the predicted departure time to be at least a threshold amount of time after a last request for access, or adjust the predicted departure in any other suitable manner.

Any suitable distribution or set of distributions may be used in embodiments that implement a distribution in this manner. For example, in some embodiments a triangular distribution may be used. A triangular distribution may be a probability distribution having a shape of an isosceles triangle. Triangular distributions are defined by a minimum, a maximum, and a mode, and as such the triangular distribution of some embodiments may have a minimum value for a number of hours spent in a space, a maximum value for a number of hours spent in a space, and mode for a number of hours. The minimum, maximum, and mode may be set to any suitable values, including set by an administrator or other user. As one example, a triangular distribution that anticipates a person working for approximately 8 hours may be used with a minimum value of 7.5 hours, a mode of 8 hours, and a maximum of 8.5 hours. With a triangular distribution having a minimum X, a maximum Y, and a mode Z, a number of hours an occupant spends in a space may be determined using a randomly-selected number R and the equations:

if (R < 0.5), hours = X + JR * (Y - X) * (Z - X)

if (R≥ 0.5), hours = Y - (\ - R) * (Y - X) * (Y - Z)

As another example, a Gaussian distribution may be used. A Gaussian distribution (also called a normal or bell-curve distribution) is defined by a mean value and a standard deviation. In embodiments that use Gaussian distributions, such values may be set to any suitable value. As one example, a Gaussian distribution that anticipates a person working for approximately 8 hours may be used with a mean value of 8 hours and a standard deviation of 3 hours. Where a Gaussian distribution is used, a similar process for determining a random value, querying the distribution based on the random value, and calculating a number of work hours based on the result from the distribution may be carried out.

As a last example, a Gaussian mixture model may be used that may account for different occupancy styles for different groups of occupants. For example, full-time employees may spend more time in a space than part-time employees. Accordingly, in some embodiments a Gaussian distribution may be determined for each of multiple different types of employees and a Gaussian mixture model determined based on those distributions. A work time for any occupant may then be determined from the mixture model without needing to determine a type of that particular occupant. The definitions of the distributions (e.g., mean and mode values, etc.) may be set in any suitable manner, as embodiments are not limited in this respect. As mentioned above, in some embodiments an administrator may set the values based on any suitable factors. The administrator may, for example, monitor behavior of occupants in a space to determine such minimum, maximum, and average work times (or other suitable values) and set the values based on that monitoring.

As another example, a utilization analysis facility may configure a distribution to be used in predicting departures of occupants from a space based on characteristics of occupants determined by the facility from an analysis of occupants in one or more other spaces. In scenarios in which occupants may provide explicit indicators of departures from a space (as discussed in more detail below), a utilization analysis facility may analyze the information about arrivals and departures in one or more spaces to determine the values. The facility may then configure a distribution according to those values and use the distribution for other spaces in which explicit indicators of departures are unavailable. As an example of using such explicit indicators of departures, in some embodiments a Gaussian distribution may be defined having mean and standard deviation values (e.g., 8 hours and 3 hours, respectively) selected by a user and this Gaussian may be modified based on explicit indicators of departures for a space over a time period. The departure data for time intervals within the time period may be represented using a histogram and a histogram equalization method may be used to transform the original Gaussian distribution based on the departure data. Because the departure data histogram is often not smooth, a second filtering step may be performed on the transformed Gaussian distribution, such as by applying a median filter to the transformed distribution.

In other cases (discussed in more detail below in connection with FIG. 15), the facility may use information regarding electronic communication signals detected in one or more spaces to analyze how occupants move in the space, including arrive and depart from the space, and may define a distribution based on values determined from that analysis. Subsequently, the facility may use the distribution to predict departure times of occupants from other spaces.

Once the utilization analysis facility determines a set of arrivals in spaces for occupants from the data received in block 1202 and infers departure times for those occupants in block 1204, the facility may in block 1206 begin analyzing utilization of spaces. The facility may determine the utilization in any suitable manner, including according to techniques similar to those discussed in connection with FIGs. 7 and 9 above. For example, the facility may determine utilization of spaces and sub-spaces during time intervals within a period of interest as well as across an entirety of a period of interest. The facility may also determine that an occupant was present in a space for a portion of a time interval or an entirety of a time interval, and determine utilization accordingly.

The facility may primarily rely on the requests for access made by a particular occupant in determining which space(s) that occupant occupied at any particular time. Specifically, the facility may assume that after a request for access is made to a space by an occupant, that occupant remained in that space until either the occupant requested access to another space or until the inferred departure time for that occupant. With each receipt of a request for access to a space (or sub-space) the occupant may be identified as being present in that space, and no longer present in the previous space. Based on such considerations, over time the utilization analysis facility may identify a presence of occupants in various spaces or sub-spaces over time, and determine utilization based on that presence and the capacity of the spaces in which they are present, as discussed above in connection with FIGs. 7 and 9.

Once the utilization is determined for each of the various spaces identified in the access requests, the process 1200 ends. Following the process 1200, the utilization information may be stored in any suitable data store and may be output to a user in any suitable manner.

While not explicitly illustrated in FIG. 12, in some embodiments the information analyzed in process 1200 (i.e., information indicative of presence, including requests for access, as well as inferred departure times) may additionally be used to classify movement behaviors of occupants and classify the effect occupants have on utilization. As discussed above in connection with FIG. 11 , in some embodiments a utilization analysis facility may maintain various classes of behavior for occupants, each of which indicates an effect that occupant has on utilization. For example, an occupant that seldom spends time in a space may be classified differently from an occupant that spends a large amount of time in the space. The process discussed above in connection with FIG. 11 used movements as indicated by electronic communication signals to determine occupant behaviors and to perform classification of occupants. It should be appreciated that a similar process may be performed based on requests for access and inferred departure times. From requests for access and inferred departure times, a utilization analysis facility may determine amounts of time an occupant spends in a space and how an occupant moves within a space, and other behaviors of the occupant. These behaviors may be compared to the conditions defining each behavior classification and used to determine an appropriate classification for an occupant. Though, it should be appreciated that embodiments are not limited to performing any classification process.

It should additionally be appreciated that while some embodiments may perform a process such as the one discussed above in connection with FIG. 12, in which a departure time from a space is inferred for each occupant, embodiments are not limited to making such an inference for each occupant. In some scenarios, in addition to requesting access to a space, individuals are required to request departure from a space or at least inform a security system of a departure. In such cases, rather than infer a departure from a space, the utilization analysis facility may use the explicit indicators of departures from the space in a determination of utilization, and in classification of an occupant according to behavior.

In the example of FIG. 12, a utilization analysis facility predicted departure times from a space for occupants using the same technique for all occupants. In some embodiments, using the same technique for all occupants, such as using the same distribution for calculation of a length of time, may not be effective. In examples given above in FIG. 12, the distributions were configured with work lengths of approximately 8 hours. If that same distribution is used for all occupants regardless of initial arrival time, then the facility may predict that an occupant who first arrived at 9 am will work for about 8 hours and that an occupant who first arrived at 1 pm will work for about 8 hours. Of course, in some spaces (e.g., professional offices) it is unlikely that an occupant who first arrived at 1 pm will work for a full 8 hours, but more likely that the occupant will work for a half day and leave at around the standard time of 5 pm. Accordingly, in some embodiments a utilization analysis facility may categorize occupants into different categories based on times associated with requests for access (or times associated with other types of information indicative of presence or arrivals) and predict a departure time for those occupants based in part on the categorization. In some such embodiments, each category may be associated with a different distribution, and a departure time for an occupant may be determined based on the distribution associated with that occupant's categorization.

FIG. 13 illustrates an example of such a process for determining departure times based on categorization of occupants and for determining utilization of spaces based on the departure times. The process 1300 of FIG. 13 has several similarities with the process 1200 of FIG. 12 and, for the sake of efficiency of description, the description of the process 1300 will reference FIG. 12 and focus on the differences with the embodiment of FIG. 12.

The process 1300 begins in block 1302, in which a utilization analysis facility receives information indicating a set of requests for access to a space. The information that is received may be of the same types of information as discussed above in connection with block 1202 of FIG. 12.

In block 1304, the facility reviews the set of requests for access to a space and, for each of one or more time periods (e.g., each of one or more days) categorizes an occupant based on the first request for access made by that occupant during that time period. The categorization is performed in block 1304 based on a time period because the categorization may vary between time periods. For example, an occupant may make a first request for access at one time on one day and make a first request for access at another time on another day, due to variations in that occupant's schedule. Making the categorization vary between time periods therefore allows for natural variations in an occupant's schedule. Any suitable time period may be used, as embodiments are not limited in this respect.

The categorization may be made by the facility based on conditions associated with each category. The conditions may be associated with any suitable information regarding a request for access, including based on time of access. For example, in some embodiments each category may be associated with a condition that a first request for access be made during a certain time interval of a time period. As a specific example, a category may be associated with a condition that a first request for access be made within one hour of the start of a work day (e.g., 9 am) on a given day.

In block 1306, the utilization analysis facility predicts a departure time for each occupant in each time period based on the categorization and the requests for access made in that time period. The facility may make the prediction in any suitable manner, and may make the prediction using techniques similar to those discussed above in connection with block 1204 of FIG. 12, such as using a random value to query a distribution. A difference between the operations of block 1204 and block 1306, though, is that the prediction may vary based on categorization, as the distribution that is used may vary between categories.

Any suitable categories may be used, as embodiments are not limited in this respect. In some embodiments, the categories may be normal working hours, late arrival, and after hours. A normal working hours category may be associated with a condition that an occupant make a first request for access within one hour of the start of a work day on a given day, while a late arrival category may be associated with a condition that a first request for access on a given day be made after the "normal working hours" interval, but before the end of the work day (e.g., 5 pm) that day. The after hours category may be associated with a condition that a first request for access on a given day be made after the end of the work day but before midnight. Each of these categories may also be associated with distributions that relate to how workers categorized into those categories may behave. For example, the normal working hours category may be associated with a distribution that assumes that occupants will depart on average after approximately 8 hours. The late arrival category, on the other hand, may be associated with either a fixed distribution that assumes that occupants will depart on average after approximately a fixed amount of time (e.g., 5 hours). In other cases, the late arrival category may not be associated with a fixed distribution, but instead a facility may generate a tailored distribution for each occupant that is categorized in that category. The tailored distribution may be based on a fixed time that occupants may be assumed to depart by, such as the end of the work day (e.g., 5 pm), with which the utilization analysis facility may be configured. When an occupant is categorized into the category, the facility may respond by determining a remaining amount of time until that fixed time, and generating a distribution based that remaining amount of time. The tailored distribution may, for example, use that remaining amount of time as a mean, median, or mode for the distribution, or use the remaining amount of time in any other manner. The after hours category may be associated with a fixed, uniform distribution that assumes an occupant will depart by a fixed time, such as midnight, without having a specific average number of hours or other value built in to the distribution. It should be appreciated, though, that these categories and distributions are merely examples, and that other examples are possible.

Once the departure time is determined for each occupant in block 1306, the utilization analysis facility may in blocks 1308 and 1310 identify movements between spaces based on requests for access and determine utilization of spaces based on the movements. The processing of blocks 1308 and 1310 may be carried out in any suitable manner, including using techniques discussed above in connection with block 1206 of FIG. 12.

Techniques have been described for determining utilization based on an analysis of electronic communication signals, and for determining utilization based on an analysis of information indicative of presence of occupants in a space through inferring departures of the occupants from the space. Above, these techniques were discussed separately. It should be appreciated, however, that in some spaces both sources of data may be available. For example, an organization that operates a space may be able to provide to a utilization analysis facility both information on electronic communication signals detected in the space as well as requests for access made for the space (e.g., requests for physical access to the space or requests for electronic access made from within the space) by occupants. In some embodiments, therefore, a utilization analysis facility may analyze both sets of data in determining utilization

FIGs. 14-15 illustrate examples of processes that may be used in some embodiments to determine utilization of a space based on an analysis of requests for access to a space and electronic signals detected in the space. More specifically, the process 1400 of FIG. 14 is an example of a process for matching occupants requesting for access to a space to occupants associated with signals detected in the space, and determining utilization from an analysis of movements of the occupants, and FIG. 15 is an example of a process for inferring rates of departures from spaces of different types based on an analysis of signals detected in spaces of those types, and using those inferred rates of departures to infer a departure time for an occupant for a space of a given type.

The process 1400 of FIG. 14 begins in block 1402, in which a utilization analysis facility receives information indicating requests for access to a space as well as information indicating electronic signals detected in a space. The information that is received in block 1402 may be of the same types of information as discussed above in connection with block 304 of FIG. 3 and block 1202 of FIG. 12. As discussed above in connection with FIGs. 3 and 12, such information may indicate for each signal a time the signal was detected in the space and an identifier for the device that emitted the signal, and may indicate for each request for access a time at which the request was received and information identifying the occupant who submitted the request. The information may be received in any suitable manner, including by the facility reading the data from a data store.

In block 1404, the facility reviews the information to identify correlations between signals emitted by devices and requests for access. Specifically, the facility reviews the information to identify correlations in times at which devices emit signals in a space and times at which requests for access are received relating to that space. Such correlations may be indicative of a single occupant who is making requests for access relating to a space and is carrying a device that is emitting signals detected in the space. This is because as an occupant approaches a space to make a request for access, or when the occupant is in a space and makes a request for access, a device carried by that occupant may emit signals that are detected in the space. A similarity in times of signals from a device and requests for access may indicate a relationship between those requests and those signals, and may therefore indicate a relationship between an occupant who made the request for access and the occupant that operates the device.

Accordingly, in block 1404 the facility may review the information about the signals and the requests for access to identify, over a period of time, correlations in times signals were detected and times requests for access were made. The facility may identify the correlations in any suitable manner, as embodiments are not limited in this respect. In some embodiments, the facility may analyze the information over a first, longer time period by analyzing portions of the information for multiple different second, shorter time periods within the first time period. As an example, the facility may analyze information about signals and requests for access over the course of a month by analyzing relationships between signals and requests in individual days.

The facility may, for example, examine the information to look for repeated correlations in first arrivals to a space by occupants, as indicated by the signals detected in the space and by the requests for access for the space. More specifically, the facility may identify for each device the first signal detected in the space over multiple days and the time at which that first signal was detected each day, which is indicative of a first arrival of the occupant operating that device to that space each day. The facility may then identify for each set of credentials presented in requests for access a first request for access made for the space each day, which is indicative of a first arrival of the occupant associated with those credentials to that space each day. The facility may then examine the times of first arrivals to look for repeated similarities. The facility may determine the repeated similarities in any suitable manner, including by determining whether a first arrival of an occupant to a space as indicated by signals emitted by a device is within a threshold amount of time of a first arrival of an occupant to that space as indicated by requests for access for that space. The facility may maintain a count of occupants that arrived within a threshold amount of time of one another on successive occasions and, when the count exceeds a threshold, identify that the occupants may be the same occupant.

The devices for which signals are analyzed in embodiments that implement a process like the process of FIG. 14 may be any suitable set of devices. In some embodiments, devices of the set may be uniquely associated (among the devices of the set) with an occupant, as identified using a process such as the one discussed above in connection with FIG. 6. In other cases, the devices may be ones that each appear to be operated by an occupant and that may be clustered into sets of one, two, or more devices per occupant such as through a process like the one discussed above in connection with FIG. 5. In the case that an occupant may be associated with multiple devices, a first arrival of the occupant in a time period may be identified as a first signal detected in the space during the time period from any of the devices in the cluster associated with that occupant.

In some embodiments, a utilization analysis facility may carry out a more detailed analysis of requests for access (e.g., requests for physical access) and electronic communication signals to determine corresponding occupants. The more detailed analysis may, for example, use statistical analysis to identify a pair of an occupant making requests for access and an occupant operating one or more devices emitting signals that is, among the potential pairs, most likely to be a correct match and/or least likely to be a coincidence.

For example, in one embodiment, the utilization analysis facility may identify different sets of devices that are operated by the same occupants, such as using techniques described above in connection with FIG. 5. This number of "clusters" of devices and, thereby, identified occupants may be larger than a number of occupants determined from analyzing requests for access for spaces. The utilization analysis facility may next examine sets of requests for access for spaces and information regarding a set of signals detected in the spaces. Specifically, the facility may examine in the sets requests and signals indicative of arrival of occupants in one or more spaces. For example, the facility may examine a first signal detected for a device in each space, which could indicate an arrival of a device (and its occupant) in that space. The facility may identify the arrivals by identifying the first requests/signals in the set for a time period, which may be a day or any other suitable period of time. The facility may then create three different three-dimensional matrices for every combination of a specific credential used in a request for access and a device. Two dimensions of the matrices are identifiers for the credential and the device and a third dimension of the matrix is an identifier for a time period (e.g., for a day). Each matrix may be associated with a time interval. In this example, three intervals (e.g., one, five, and ten minutes) may be used, but it should be appreciated that any suitable number of intervals each having any suitable length may be used, as embodiments are not limited in this respect. The facility may initialize all cells of the matrices to values of 0 and then add a value of 1 to a cell of a matrix when the information about the requests and signals indicate that the credential and the device arrived in one of the spaces within the time interval (one, five, or ten minutes) for that matrix. The facility may then sum each of the three-dimensional matrices along the time axis to determine three corresponding two-dimensional matrices, which each indicate a number of times a particular device and a particular credential arrived in a space close in time to one another. Again, each of the three two-dimensional matrices corresponds to a time interval (e.g., one, five, or ten minutes). Then, for each credential in each matrix, the facility may identify a device having a highest value in that matrix and identify that credential-device pair as a potential match.

Next, the facility may continue to analyze potential combinations of credentials and devices as indicated by the information indicating a set of requests for access for spaces and a set of signals detected in the spaces. The facility may calculate from the information multiple different scores, each of which are two-dimensional matrix with one dimension being identifiers for credentials used in requests for access and another dimension being identifiers for clusters of devices (e.g., as previously determined using a technique such as the one discussed in connection with FIG. 5).

For the first of the scores, the facility may calculate a value of a cell based on an analysis of the credential corresponding to that cell and each of the devices of the device cluster corresponding to that cell. For each cell, the facility may first calculate an average time difference (e.g., in hours) in the times indicated for the arrival of the credential and the arrival of each device in each space, and then average the values for each of the devices to yield an average value for the cluster. A logarithm of that average value may then be calculated and added to 1 , and a reciprocal of that sum stored as the value of the cell. This process may be repeated for each of the cells (i.e., each pair of credential and cluster).

The facility may next calculate three different matrices using the matrices calculated in the first step above. As mentioned above, each of the three different two-dimensional matrices (each of which corresponds to a time interval) calculated above included one dimension corresponding to a device and another dimension corresponding to a credential. The facility may produce three matrices that each correspond to a time interval, but that include along one axis an identifier for a cluster of devices. For each cell of a credential-cluster matrix, the facility may analyze, in the matrix for the corresponding time interval, the row of the matrix corresponding to the credential. Specifically, the facility may analyze the row to determine how many of the devices in the cluster have a count of arrivals above a threshold. The facility may store in the cell of the matrix corresponding to the credential and the cluster the number of devices in the cluster that had the count of arrivals above the threshold. The facility may repeat this process for each credential-cluster pair for each of the three matrices, and then normalize the cells in each of the three matrices by the highest valued cell in the results.

The facility may next calculate three additional matrices in a manner similar to the three matrices in the preceding paragraph. The matrices may again have credential-cluster dimensions and correspond to the time intervals discussed above, and the facility may initially determine a value of each cell using the same process of counting devices in each cluster having a count of arrivals above a threshold, as in the preceding paragraph. However, whereas the matrices calculated as in the preceding paragraph were modified in a last step using a normalization, these three matrices are modified as a last step using Poisson distributions. For each matrix, a Poisson distribution is calculated that has as a mean value a mean of the values calculated for that matrix. Each matrix is then modified by calculating, for each cell and based on the count value stored in each cell, a density in the Poisson distribution corresponding to the count value.

The facility may next calculate another two-dimensional matrix having a dimension corresponding to credentials and a dimension corresponding to clusters of devices. Each cell of the matrix may be determined as a count of a number of time periods (e.g., a number of days) in which a request for access was made for a space and a signal emitted by a device in the cluster was detected in the space. This is a count of a number of time periods that both the credential and the devices of the cluster were used in the space. Once the count is determined for each cell of the matrix, the matrix is normalized using a sigmoid function. Subsequently, a median of the values is calculated and each cell of the matrix is further modified by subtracting the median and dividing by three.

Using the techniques of the four preceding paragraphs, four matrices are determined for each time interval: two constant matrices from the first and fourth of these paragraphs, and one matrix for each of the three time intervals from the second and third of these paragraphs. Next, the facility may calculate a weighted sum matrix for each time interval by, for each time interval, multiplying each of the four matrices by a corresponding weighting value and then summing corresponding cells in the weighted matrices. Then, for each of the three matrices, the facility may determine an optimal pairing of credential to cluster (and thus, an optimal pairing of credential occupant to device cluster occupant) using the Hungarian algorithm. The facility may calculate the weighted sum matrix for each of multiple possible sets of weights for the four matrices. Through this process, the facility may calculate multiple different weighted sum matrices for each time interval and may, for each weighted sum matrix, determine an optimal pairing using the Hungarian algorithm.

Next, the facility may, for each of the three time intervals, find a most stable matching of credential to device cluster from each of the multiple weighted sum matrices (as calculated in the preceding paragraph) calculated for that time interval. The facility may generate a matrix for each time interval with index of matching as index of columns and index of rows.

The facility may next count a number of different matches that are near the matching resulted from each set of weights, and add the vector to the row of the matrix according to the set of weights. The facility may then find a most stable matching, but in this case through picking a matching that has a highest diagonal value in the matrix.

Next, the facility may analyze three different matrices that result for each of the three time intervals to identify, for each device cluster, whether the matrices indicate a potential pairing for that device cluster. Specifically, the facility may first analyze the most stable matching identified for each device cluster for each time interval and, if the most stable matching for a device cluster is the same credential for all three time intervals, identify that pair of device cluster and credential as a match. Next, for the remaining unmatched device clusters, the facility may analyze the three matrices to identify, for each device cluster, whether a particular credential was identified as the most stable match for the cluster for two out of the three time intervals. If so, that pair of device cluster and credential are identified as a match. Next, for the remaining unmatched device clusters, the facility analyzes the three matrices to identify, for each device cluster, whether a credential was identified as a stable match for the cluster for one of the time intervals. If so, that pair of device cluster and credential is stored as a match.

Using this more detailed process, each device cluster and its occupant may be matched to a credential and its occupant, thereby identifying occupants that use one or more devices of a cluster and use a credential to make requests for access.

Once the facility has identified correlations between signals detected in a space and requests for access for that space, the facility may match occupants that operate the devices and occupant that made the requests for access and identify them as the same occupant.

Accordingly, if the facility identifies that a device is correlated with requests for access, the facility may identify that a single occupant both operates that device and made those requests for access. When the facility matches occupants in this manner, the facility may store information identifying that the device and the requests were made by a same occupant. Having matched devices to requests for access in this way, the utilization analysis facility may be able to provide a very fine grain of information about utilization of a space. Signals may by themselves allow for a fine grain of utilization information to be determined, as the devices that emitted the signals can be localized to a specific position and movements of those devices (and the occupants that operate them) can be tracked closely using the signals, and the analysis improves as the device emits more signals. However, as should be appreciated from the foregoing, it may be difficult to identify a specific occupant operating a device and, as such, it may be difficult to determine any information about the occupant operating the device other than the person exists (hence the discussion above of occupants being inferred persons). Requests for access can provide a large amount of information about a (in some embodiments, anonymous) person making the request, because the requests for access may include credential information specifically associated with an individual. With that credential, demographic information, occupation information, and other information about that individual can be retrieved. Though, requests for access may not provide as much information about movements of an individual as an analysis of signal data can provide.

By matching occupants who operate devices to occupants who make requests for access, however, the advantages of both types of data can be realized. Once an occupant who operates one or more devices has been identified as the individual who presented credentials in a request for access, the identifying information, demographic information, occupation information, etc. for the individual can be associated with the device(s), and the movements indicated by the device(s) can be associated with that information.

Accordingly, in block 1406, the utilization analysis facility tracks movements of the matched occupants in a space based on an analysis of signals emitted by devices operated by each of the matched occupants. This tracking may be performed in any suitable manner, including according to techniques discussed above in connection with FIGs. 7-8. In some embodiments, while the matching of occupants in block 1404 may have been done using multiple devices of a cluster associated with an occupant, the utilization analysis facility may perform the tracking of block 1406 using just one device per occupant, for reasons that should be appreciated from the foregoing. In some embodiments, in addition to tracking matched occupants, the facility may additionally track unmatched occupants using the techniques discussed above.

In block 1408, the utilization analysis facility determines utilization based on the movements determined from signal analysis in block 1406. The determination may be made in any suitable manner, such as using techniques discussed above. However, in addition to determining utilization for all occupants, the facility may determine utilization for different classes of matched occupants. The classes may be based on characteristics of individuals and each may be based on any suitable characteristic(s) for differentiating between groups of individuals. For example, characteristics that may be determined from the demographic information or occupation information associated with requests for access may be used, such that the classes may be based on demographic characteristics or occupation characteristics. As specific examples, classes based on gender, age, job title, job department, or other characteristics may be used. When determining utilization based on the classes, the facility may use the same techniques for determining utilization (e.g., presence of occupants as compared to capacity of a space) as discussed above, but may determine the utilization for each class based on the presence of occupants of that class in a space. For example, if the facility detects the presence of 10 occupants in a space during a time period, but only two of them fall within a certain class, to determine the utilization of that space for that class during the time period the facility will only consider two occupants to have been present.

The facility may determine a class of an occupant based on the information determined about the occupant from the matching discussed above. The information requests for access may include (or enable retrieval of) identifying information, demographic information, occupation information, and/or other information about an occupant that made the request for access. Once that occupant is matched to a device using the foregoing technique, that information may be used to characterize the occupant that operates the device and used in determining whether the occupant satisfies the characteristics associated with one or more classes.

Once the utilization is determined for a space as a whole and/or for each of one or more classes of occupant, the utilization and class-based utilization(s) may be stored in any suitable data store and/or may be output to a user in any suitable manner. After the determined utilization(s) are stored, the process 1400 ends.

A technique for using a distribution of lengths of time to make a probabilistic prediction of a length of time an occupant will remain in a space following arrival in the space, and thereby a departure time of the occupant from the space, was discussed above in connection with FIG. 12. In that discussion, the distribution was described as being determined in any of various ways including based on user input or based on an analysis of departures from other spaces as determined by explicit requests to leave that are available for those other spaces or from an analysis of departures from other spaces as indicated by electronic communication signals detected in those other spaces. FIG. 15 illustrates an example of a process that may be used in some embodiments for determining a distribution of lengths of time that may be used to infer departures from a space. The process 1500 of FIG. 15 begins in block 1502, in which a utilization analysis facility determines arrivals and departures for various spaces using signals detected in those spaces. As part of determining the arrivals and departures for a space, the facility may determine a length of time that each occupant spends within that space. The analysis of the signals may be performed in any suitable manner, including according to techniques described above.

In block 1504, the facility may additionally categorize spaces. The facility may perform the categorization in any suitable manner. In some embodiments, the facility may receive from a user an input of a categorization of each of the spaces according to a set of categories. Any suitable categories may be used, as embodiments are not limited in this respect. In other embodiments, the facility may use a machine learning and clustering technique to determine categorizations itself. In such an embodiment, the facility may review timing characteristics for arrivals and departures from each of various spaces and cluster spaces together based on those timing characteristics. For example, if two spaces exhibit a similarity in arrival times and departure times by occupants, the spaces may be clustered together. The facility may then set the categories of spaces based on the clustering. Known clustering algorithms may be used in such embodiments. Those skilled in the art will understand how to implement such a clustering algorithm based on these inputs and goals.

In block 1506, once arrivals and departures, and lengths of time, have been determined for various spaces and those spaces have been categorized, the facility may determine a distribution of lengths of time for each of the categories. The distribution for each category may be determined from the lengths of time in any suitable manner, including using counts of observed lengths of time to determine probabilities associated with each length of time and determining the distribution using those probabilities. Where such counts are used, in some embodiments the facility may aggregate individual lengths of time into a set of intervals of time to simplify the analysis. For example, where the facility may observe occupants staying in a space for 16 minutes and 17.5 minutes, the facility may simply analyze those as two spans of 15- 20 minutes. Any suitable set of time intervals may be used, as embodiments are not limited in this respect.

In some embodiments, in addition to determining a distribution of lengths of time for each category, the facility may determine characteristics of arrivals for each category. The characteristics of arrivals may be described in any suitable manner, including using a distribution that describes sequences of events (e.g., a Poisson distribution). In such embodiments, the characterization of the arrivals for each category may be used to determine a proper

categorization of a space based on observed arrivals in that space, through comparing the observed arrivals to the characterization for each category and selecting a best match. Once the distributions for the categories are determined, the distributions may be used to predict departures from spaces by occupants. In block 1508, the utilization analysis facility receives as input a set of requests for access for a space (or other information indicative of presence of occupants, including arrivals of occupants to a space). Such requests for access may be requests for physical or electronic access, and may include any suitable information, including the information described in connection with block 1202 of FIG. 12. In block 1508, the facility may also determine a categorization of the space to which the requests relate. The facility may determine that categorization in any suitable manner, including by receiving explicit input from a user or by analyzing timing characteristics of the requests and identifying a best match to timing characteristics for each of the categories, as discussed above.

Once a category is determined in block 1508, using the distribution associated with that category and the information regarding the requests for access, the facility in block 1510 determines predicted departure times from the space for each of the occupants indicated by the requests for access for the space and, in block 1512, determines utilization based on the requests and the predicted departures. The analysis of blocks 1510 and 1512 may be performed in any suitable manner, including according to techniques described above in connection with FIGs. 12-13. Once the utilization is determined, the process 1500 ends.

The process 1500 was described as determining the distribution from an analysis of lengths of time as indicated by arrivals and departures for spaces inferred from electronic communication signals detected in that space. It should be appreciated that embodiments are not so limited. In some embodiments, a similar process may be carried out that uses requests for access and explicit signals for departures from a space (e.g., requests to depart) to identify lengths of time that occupants spend in a space. In still other embodiments, both lengths of time inferred from electronic communication signals and inferred from requests for access and explicit signals for departures may be used by a facility to develop a distribution. In some such embodiments that use both sources of data, a distribution may be calculated independently using both sets of data and a final distribution may be calculated for a space or a category of space using a weighted sum of the two distributions. Accordingly, in some embodiments, distributions for categories may be set based on distributions determined from analysis of electronic communication signals, analysis of requests for access and explicit signals of departures, and/or analysis of both electronic communication signals and requests for access/explicit signals of departures.

The exemplary process of FIG. 15 used information derived from an analysis of electronic communication signals to inform an analysis of utilization of a space that was based primarily on requests for access for a space. Specifically, while the requests for access for a space were used in determining arrivals to the space and a number of occupants in the space, analysis of electronic communication signals contributed a distribution that was used in the prediction of departure times for occupants.

In some embodiments, a utilization analysis that is primarily based on an analysis of electronic communication signals may be informed by information on requests for access for a space, or otherwise based on other information indicative of a presence of occupants in a space. More particularly, in some embodiments a number of occupants in a space during a time period may be separately determined based on (1) an analysis of electronic communication signals detected in the space during the time period, and (2) an analysis of other information indicative of presence of occupants in a space during the time period, such as requests for access for a space. The two numbers for the number of occupants in the space may be used to determine utilization.

For example, as should be appreciated from the foregoing, information on electronic communication signals detected in the space may provide for a more detailed indication of utilization, as it may provide more detailed information on how occupants move during a time period. Though, based on the uncertainties of algorithmic analysis of signals discussed above, a number of occupants determined using the signal analysis may be less precise than a number of occupants that may be determined using the other types of information, such as requests for access for a space.

Accordingly, in some embodiments a number of occupants that were present in a space over a time period may be determined using both sets of information, and a ratio determined. The ratio may be, for example, a number of occupants determined from an analysis of information indicative of presence of occupants to a number of occupants determined from analysis of electronic communication signals. Based on perceived reliability of the number of occupants determined from the analysis of information indicative of presence (e.g., requests for access), the ratio is indicative of an estimated error in the analysis of utilization of that space over that time period based on electronic communication signals alone.

Subsequently, utilization of a space (and, in some cases, of sub-spaces) over a time period may be first determined based on the analysis of electronic communication signals using techniques discussed above, which may include determining a number of occupants in a space or a sub-space and utilization of the space and sub-space over time intervals. Once the utilization is determined, the utilization may be adjusted based on the estimated error. For example, each utilization determination for a space or sub-space over the time period or a time interval may be adjusted based on the estimated error. The adjustment may be made in any suitable manner. As should be appreciated from the foregoing, a calculation of utilization may involve a comparison of a number of occupants in a space at a time to a capacity of a space. In some embodiments, the adjustment may be made to each calculation of utilization by multiplying the number of occupants used in that calculation by the estimated error (e.g., multiplying by the ratio discussed above).

Techniques operating according to the principles described herein may be implemented in any suitable manner. Included in the discussion above are a series of flow charts showing the steps and acts of various processes that estimate utilization of a space (and sub-spaces of the space) based on analyzing electronic signals detected in the space and information indicating the presence of occupants in the space, such as information indicating requests for access for the space (e.g., requests for physical access to the space or requests for electronic access received from within the space). The processing and decision blocks of the flow charts above represent steps and acts that may be included in algorithms that carry out these various processes.

Algorithms derived from these processes may be implemented as software integrated with and directing the operation of one or more single- or multi-purpose processors, may be implemented as functionally-equivalent circuits such as a Digital Signal Processing (DSP) circuit or an

Application-Specific Integrated Circuit (ASIC), or may be implemented in any other suitable manner. It should be appreciated that the flow charts included herein do not depict the syntax or operation of any particular circuit or of any particular programming language or type of programming language. Rather, the flow charts illustrate the functional information one skilled in the art may use to fabricate circuits or to implement computer software algorithms to perform the processing of a particular apparatus carrying out the types of techniques described herein. It should also be appreciated that, unless otherwise indicated herein, the particular sequence of steps and/or acts described in each flow chart is merely illustrative of the algorithms that may be implemented and can be varied in implementations and embodiments of the principles described herein.

Accordingly, in some embodiments, the techniques described herein may be embodied in computer-executable instructions implemented as software, including as application software, system software, firmware, middleware, embedded code, or any other suitable type of computer code. Such computer-executable instructions may be written using any of a number of suitable programming languages and/or programming or scripting tools, and also may be compiled as executable machine language code or intermediate code that is executed on a framework or virtual machine.

When techniques described herein are embodied as computer-executable instructions, these computer-executable instructions may be implemented in any suitable manner, including as a number of functional facilities, each providing one or more operations to complete execution of algorithms operating according to these techniques. A "functional facility," however instantiated, is a structural component of a computer system that, when integrated with and executed by one or more computers, causes the one or more computers to perform a specific operational role. A functional facility may be a portion of or an entire software element. For example, a functional facility may be implemented as a function of a process, or as a discrete process, or as any other suitable unit of processing. If techniques described herein are implemented as multiple functional facilities, each functional facility may be implemented in its own way; all need not be implemented the same way. Additionally, these functional facilities may be executed in parallel and/or serially, as appropriate, and may pass information between one another using a shared memory on the computer(s) on which they are executing, using a message passing protocol, or in any other suitable way.

Generally, functional facilities include routines, programs, objects, components, data structures, etc. that perform particular tasks or implement particular abstract data types.

Typically, the functionality of the functional facilities may be combined or distributed as desired in the systems in which they operate. In some implementations, one or more functional facilities carrying out techniques herein may together form a complete software package. These functional facilities may, in alternative embodiments, be adapted to interact with other, unrelated functional facilities and/or processes, to implement a software program application.

Some exemplary functional facilities have been described herein for carrying out one or more tasks. It should be appreciated, though, that the functional facilities and division of tasks described is merely illustrative of the type of functional facilities that may implement the exemplary techniques described herein, and that embodiments are not limited to being implemented in any specific number, division, or type of functional facilities. In some implementations, all functionality may be implemented in a single functional facility. It should also be appreciated that, in some implementations, some of the functional facilities described herein may be implemented together with or separately from others (i.e., as a single unit or separate units), or some of these functional facilities may not be implemented.

Computer-executable instructions implementing the techniques described herein (when implemented as one or more functional facilities or in any other manner) may, in some embodiments, be encoded on one or more computer-readable media to provide functionality to the media. Computer-readable media include magnetic media such as a hard disk drive, optical media such as a Compact Disk (CD) or a Digital Versatile Disk (DVD), a persistent or non- persistent solid-state memory (e.g., Flash memory, Magnetic RAM, etc.), or any other suitable storage media. Such a computer-readable medium may be implemented in any suitable manner, including as computer-readable storage media 1606 of FIG. 16 described below (i.e., as a portion of a computing device 1600) or as a stand-alone, separate storage medium. As used herein, "computer-readable media" (also called "computer-readable storage media") refers to tangible storage media. Tangible storage media are non-transitory and have at least one physical, structural component. In a "computer-readable medium," as used herein, at least one physical, structural component has at least one physical property that may be altered in some way during a process of creating the medium with embedded information, a process of recording information thereon, or any other process of encoding the medium with information. For example, a magnetization state of a portion of a physical structure of a computer-readable medium may be altered during a recording process.

In some, but not all, implementations in which the techniques may be embodied as computer-executable instructions, these instructions may be executed on one or more suitable computing device(s) operating in any suitable computer system, including the exemplary computer system of FIG. 16, or one or more computing devices (or one or more processors of one or more computing devices) may be programmed to execute the computer-executable instructions. A computing device or processor may be programmed to execute instructions when the instructions are stored in a manner accessible to the computing device or processor, such as in a data store (e.g., an on-chip cache or instruction register, a computer-readable storage medium accessible via a bus, a computer-readable storage medium accessible via one or more networks and accessible by the device/processor, etc.). Functional facilities comprising these computer-executable instructions may be integrated with and direct the operation of a single multi-purpose programmable digital computing device, a coordinated system of two or more multi-purpose computing device sharing processing power and jointly carrying out the techniques described herein, a single computing device or coordinated system of computing device (co-located or geographically distributed) dedicated to executing the techniques described herein, one or more Field-Programmable Gate Arrays (FPGAs) for carrying out the techniques described herein, or any other suitable system.

FIG. 16 illustrates one exemplary implementation of a computing device in the form of a computing device 1600 that may be used in a system implementing techniques described herein, although others are possible. It should be appreciated that FIG. 16 is intended neither to be a depiction of necessary components for a computing device to operate as a computing device in accordance with the principles described herein, nor a comprehensive depiction.

Computing device 1600 may comprise at least one processor 1602, a network adapter 1604, and computer-readable storage media 1606. Computing device 1600 may be, for example, a desktop or laptop personal computer, a personal digital assistant (PDA), a mobile phone, a server, a wireless access point or other networking element, or any other suitable computing device. Network adapter 1604 may be any suitable hardware and/or software to enable the computing device 1600 to communicate wired and/or wirelessly with any other suitable computing device over any suitable computing network. The computing network may include wireless access points, switches, routers, gateways, and/or other networking equipment as well as any suitable wired and/or wireless communication medium or media for exchanging data between two or more computers, including the Internet. Computer-readable media 1606 may be adapted to store data to be processed and/or instructions to be executed by processor 1602. Processor 1602 enables processing of data and execution of instructions. The data and instructions may be stored on the computer-readable storage media 1606 and may, for example, enable communication between components of the computing device 1600.

The data and instructions stored on computer-readable storage media 1606 may comprise computer-executable instructions implementing techniques which operate according to the principles described herein. In the example of FIG. 16, computer-readable storage media 1606 stores computer-executable instructions implementing various facilities and storing various information as described above. Computer-readable storage media 1606 may store a utilization analysis facility 1608 implementing any or all of the techniques described above. The storage media 1606 may additionally store data 1610 that includes information on electronic

communication signals and/or requests for access, which may include any of the illustrative types of information discussed above. Data 1612 identifying blacklisted devices may be stored in some embodiments, as may data 1614 that identifies information about spaces, such as floor plans, capacities, coordinates for the space and for sub-spaces, relationships between credential- reading devices and spaces, or other suitable information. Storage media 1616 may also store utilization data 1616 that may have been produced by the utilization analysis facility 1608.

While not illustrated in FIG. 16, a computing device may additionally have one or more components and peripherals, including input and output devices. These devices can be used, among other things, to present a user interface. Examples of output devices that can be used to provide a user interface include printers or display screens for visual presentation of output and speakers or other sound generating devices for audible presentation of output. Examples of input devices that can be used for a user interface include keyboards, and pointing devices, such as mice, touch pads, and digitizing tablets. As another example, a computing device may receive input information through speech recognition or in other audible format.

Embodiments have been described where the techniques are implemented in circuitry and/or computer-executable instructions. It should be appreciated that some embodiments may be in the form of a method, of which at least one example has been provided. The acts performed as part of the method may be ordered in any suitable way. Accordingly, embodiments may be constructed in which acts are performed in an order different than illustrated, which may include performing some acts simultaneously, even though shown as sequential acts in illustrative embodiments.

Various aspects of the embodiments described above may be used alone, in combination, or in a variety of arrangements not specifically discussed in the embodiments described in the foregoing and is therefore not limited in its application to the details and arrangement of components set forth in the foregoing description or illustrated in the drawings. For example, aspects described in one embodiment may be combined in any manner with aspects described in other embodiments.

Use of ordinal terms such as "first," "second," "third," etc., in the claims to modify a claim element does not by itself connote any priority, precedence, or order of one claim element over another or the temporal order in which acts of a method are performed, but are used merely as labels to distinguish one claim element having a certain name from another element having a same name (but for use of the ordinal term) to distinguish the claim elements.

Also, the phraseology and terminology used herein is for the purpose of description and should not be regarded as limiting. The use of "including," "comprising," "having,"

"containing," "involving," and variations thereof herein, is meant to encompass the items listed thereafter and equivalents thereof as well as additional items.

The word "exemplary" is used herein to mean serving as an example, instance, or illustration. Any embodiment, implementation, process, feature, etc. described herein as exemplary should therefore be understood to be an illustrative example and should not be understood to be a preferred or advantageous example unless otherwise indicated.

Having thus described several aspects of at least one embodiment, it is to be appreciated that various alterations, modifications, and improvements will readily occur to those skilled in the art. Such alterations, modifications, and improvements are intended to be part of this disclosure, and are intended to be within the spirit and scope of the principles described herein. Accordingly, the foregoing description and drawings are by way of example only.

Claims

CLAIMS What is claimed is:

1. A method for tracking utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space, the method comprising:

operating at least one programmed processor to perform acts comprising:

receiving information indicating at least a first set of electronic communication signals detected in the space over a first time period, the first set of electronic communication signals having been emitted by a first set of devices, the information indicating at least, for each detected electronic communication signal, an identifier for one of the first set of devices that emitted that detected electronic communication signal, an estimated location of the one device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected;

determining, based on the information indicating the set of electronic communication signals emitted by the first set of devices, a limited set of devices that are each uniquely associated, among the first set of devices, with an occupant of the space, the limited set of devices being a subset of the first set of devices, wherein the determining the limited set that are uniquely associated comprises, in a case that two or more devices of the first set of devices are determined to be associated with one occupant, identifying only one of the two or more devices to include in the limited set of devices; and

determining utilization of the space during the time period of interest based at least in part on information indicating a second set of electronic communication signals that were detected as having been emitted by devices of the limited set of devices.

2. The method of claim 1, wherein:

the first time period includes the time period of interest; and receiving the information indicating at least the first set of electronic communication signals detected in the space over the first time period comprises receiving information indicating the second set of electronic communication signals detected in the space over the time period of interest.

3. The method of claim 1, wherein determining the limited set comprises:

identifying, from an analysis of information regarding electronic communication signals of the first set of signals emitted by a first device of the first set of devices, a first behavior exhibited by the first device;

identifying, from an analysis of information regarding electronic communication signals of the first set of signals emitted by a second device of the first set of devices, a second behavior exhibited by the second device;

in response to identifying a correlation between the first behavior and the second behavior, identifying that the first device and the second device are operated by a same occupant.

4. The method of claim 3, wherein determining the limited set further comprises:

in response to identifying that the first device and the second device are operated by the same occupant, selecting only one of the first device and the second device to include in the limited set.

5. The method of claim 4, wherein selecting only one of the first device and the second device comprises selecting a more active device of the first device and the second device.

6. The method of claim 4, wherein:

selecting only one of the first device and the second device comprises selecting the first device;

determining utilization of the space during the time period of interest comprises determining the utilization of the space during a first portion of the time period of interest based at least in part on information indicating a second set of electronic communication signals that were detected as having been emitted by devices of the limited set of devices during the first portion of the time period of interest; and the method further comprises determining the utilization of the space during a second portion of the time period of interest based at least in part on information indicating a third set of electronic communication signals that were detected in the space during the second portion of the time period of interest, wherein determining the utilization of the space during the second portion comprises:

in response to determining, based on the information indicating the third set of electronic communication signals, that the third set does not include signals emitted by the first device and that the third set includes signals emitted by the second device, determining the utilization based at least in part on one or more signals that the information on the third set indicates were emitted by the second device.

7. The method of claim 3, wherein determining the limited set further comprises:

identifying, from an analysis of information regarding electronic communication signals of the first set of signals emitted by a second device of the first set of devices, a third behavior exhibited by a third device;

in response to identifying a correlation between the first behavior and the third behavior, identifying that the first device and the third device are operated by the same occupant.

8. The method of claim 1, wherein determining utilization of the space during the time period of interest based at least in part on information indicating a second set of electronic communication signals comprises determining whether a device of the limited set of devices was present in the space during the time period of interest.

9. The method of claim 8, wherein determining whether the device of the limited set of devices was present in the space during the time period of interest comprises:

reviewing the second set of electronic communication signals to determine whether the second set includes a signal emitted by the device from within the space during a period of time before the time period of interest;

reviewing the second set of electronic communication signals to determine whether the second set includes a signal emitted by the device from within the space during the time period of interest; and reviewing the second set of electronic communication signals to determine whether the second set includes a signal emitted by the device from within the space during a period of time after the time period of interest.

10. The method of claim 1, wherein determining the utilization of the space during the time period of interest comprises determining a utilization of a second space within the space during the time period of interest.

11. The method of claim 1 , wherein determining the utilization of the space during the time period of interest comprises:

determining, from an analysis of the second set of electronic communication signals, a number of occupants present in the space during the time period of interest; and

comparing the number of occupants present in the space during the time period of interest to a capacity of the space.

12. The method of claim 1, further comprising:

receiving information indicating a set of requests for physical access to the space made by occupants during the time period of interest;

determining an estimated error in the utilization based at least in part on the information indicating the set of requests for physical access; and

adjusting the utilization based at least in part on the estimated error.

13. At least one computer-readable storage medium having encoded thereon executable instructions that, when executed by at least one computer, cause the at least one computer to carry out a method for enabling tracking of utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space, the method comprising:

receiving information indicating electronic communication signals detected in the space over a first time period different from the time period of interest, the electronic communication signals detected in the space having been emitted by a first set of devices, the information indicating at least, for each detected electronic communication signal, an identifier for a device that emitted that detected electronic communication signal, an estimated location in the space of a device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected;

determining, based at least in part on the information indicating the electronic communication signals detected in the space over the first time period, that at least a first device and a second device of the first set of devices that emitted electronic communication signals indicated in the information are associated with one occupant of the space; and

evaluating utilization of the space during the time period of interest based on electronic communication signals detected in the space at least in part by filtering, from second information indicating electronic communication signals detected in the space during the time period of interest, electronic communication signals emitted by the second device.

14. The at least one computer-readable storage medium of claim 13, wherein determining that the first device and second device are associated with the one occupant comprises:

identifying, from an analysis of information regarding electronic communication signals emitted by the first device, a first behavior exhibited by the first device;

identifying, from an analysis of information regarding electronic communication signals emitted by the second device, a second behavior exhibited by the second device;

in response to identifying a correlation between the first behavior and the second behavior, identifying that the first device and the second device are associated with the one occupant.

15. The at least one computer-readable storage medium of claim 14, wherein identifying the first behavior exhibited by the first device comprises determining at least one timing

characteristic of the electronic communication signals emitted by the first device.

16. The at least one computer-readable storage medium of claim 14, wherein: the method further comprises determining, from an analysis of the electronic

communication signals emitted by the first device and by the second device, which of the first device and the second device the one occupant uses as a primary device; and

filtering the electronic communication signals emitted by the second device is performed in response to determining that the one occupant uses the first device as the primary device.

17. The at least one computer-readable storage medium of claim 16, wherein determining which of the first device and the second device the one occupant uses as the primary device comprises evaluating a device type of the first device and the second device.

18. The at least one computer-readable storage medium of claim 16, wherein determining which of the first device and the second device the one occupant uses as the primary device comprises determining, from an analysis of the electronic communication signals emitted by the first device and by the second device, which of the first device and the second device is more active.

19. An apparatus comprising:

at least one processor; and

at least one storage medium having encoded thereon executable instructions that, when executed by the at least one processor, cause the at least one processor to carry out a method for enabling tracking of utilization of a space by occupants through evaluating electronic communication signals detected in the space, the method comprising:

receiving first information indicating electronic communication signals detected in the space over a time period, the electronic communication signals detected in the space having been emitted by a first set of devices, the information indicating at least, for each detected electronic communication signal, an identifier for a device that emitted that detected electronic communication signal, an estimated location in the space of a device that emitted that detected electronic communication signal, and a time that detected electronic communication signal was detected; determining, based at least in part on the information indicating the electronic communication signals emitted by the first set of devices and detected in the space over the first time period, a first set of occupants that operate the first set of devices, a number of occupants in the first set of occupants being less than a number of devices in the first set of devices;

based at least in part on the estimated locations of the electronic communication signals indicated by the information, identifying movements of each occupant of the first set of occupants during the time period;

based on the movements, classifying each of the occupants in the first set of occupants into one of a set of mobility classifications, the mobility classifications of the set indicating an effect of an occupant on occupancy of the space.

20. The apparatus of claim 19, wherein:

determining the first set of occupants that operate the first set of devices comprises determining that a first occupant of the first set operates at least a first device and a second device of the first set of devices;

the method further comprises determining, from an analysis of the electronic

communication signals emitted by the first device and by the second device, which of the first device and the second device the first occupant uses as a primary device; and

identifying movements of each occupant of the first set of occupants comprises identifying the movements based on an analysis of electronic communication signals emitted by the determined primary device.

21. The apparatus of claim 19, wherein:

each classification of the set of mobility classifications is associated with a behavior characteristic; and

classifying a first occupant of the occupants into one of the set of mobility classifications comprises:

evaluating the movements of the first occupant, determined from the analysis of the electronic communication signals, to determine a behavior characteristic exhibited by the movements; and comparing the behavior characteristic exhibited by the first occupant to the behavior characteristics for the mobility classifications to determine whether the behavior characteristic matches one or more of the behavior characteristics for the mobility classifications.

22. At least one computer-readable storage medium having encoded thereon executable instructions that, when executed by at least one processor, cause the at least one processor to carry out a method for enabling tracking of utilization of a space by occupants during a time period of interest through evaluating electronic communication signals detected in the space, the method comprising:

identifying at least one device of the first set of devices for which emission of communication signals by the at least one device during the time period of interest is not to be considered in evaluating utilization of the space during the time period of interest, wherein identifying the at least one device comprises evaluating the information indicating the electronic communication signals detected in the space over the first time period; and

evaluating utilization of the space during the time period of interest at least in part by filtering, from second information indicating electronic communication signals detected in the space during the time period of interest, electronic communication signals emitted by the at least one device.

23. The at least one computer-readable storage medium of claim 22, wherein identifying the at least one device for which emission of communication signals by the at least one device during the time period of interest is not to be considered comprises identifying at least one network infrastructure device that emitted signals indicated by the received information.

24. The at least one computer-readable storage medium of claim 22, wherein identifying the at least one device comprises:

reviewing the received information to identify, from the information regarding a signal emitted by a first device of the first set of devices, an identifier for the first device;

determining whether the identifier indicates that the first device is of a device type that is not to be considered in evaluating utilization; and

in response to determining that the identifier indicates that the first device is of a device type that is not to be considered in evaluating utilization, identifying the first device as one of the at least one device.

25. The at least one computer-readable storage medium of claim 22, wherein identifying the at least one device comprises:

reviewing the received information to identify, from the information regarding electronic communication signals emitted by a first device of the first set of devices, a behavior of the first device;

determining whether the behavior indicates that the first device is of a device type that is not to be considered in evaluating utilization; and

in response to determining that the behavior indicates that the first device is of a device type that is not to be considered in evaluating utilization, identifying the first device as one of the at least one device.

26. A method for tracking utilization of a space by occupants during a time period of interest through evaluating requests for access for spaces, the requests having been made by occupants over time, the method comprising:

operating at least one programmed processor to perform acts comprising:

receiving information indicating a set of requests for access for a space made by a set of occupants during a time period of interest, the information indicating for each request for access of the set at least a time that the request was made; predicting, based on the received information indicating the set of requests for access, departure times for each occupant of the set of occupants, the departure times indicating times at which each occupant will leave the space during the time period of interest; and determining utilization of the space during the time period of interest based at least in part on the requests for access and the departure times.

27. The method of claim 26, wherein predicting the departure times comprises, for each occupant:

reviewing a distribution of lengths of times occupants may spend in the space to determine a length of time; and

calculating the departure time by adding the length of time to a time that the received information indicates that the occupant made the request for access.

28. The method of claim 27, wherein reviewing the distribution of lengths of time occupants may spend in the space comprises:

determining a categorization of the space among a set of space categories; and retrieving the distribution associated with the determined categorization.

29. The method of claim 27, further comprising:

analyzing electronic communication signals detected in a second space to determine lengths of time that occupants of the second space spend in the second space; and

calculating the distribution of lengths of time based on a result of the analyzing.

30. A method for tracking utilization of a space by occupants during a time period of interest by evaluating electronic communication signals detected in the space, the method comprising: operating at least one programmed processor to perform acts comprising:

receiving information indicating at least a first set of electronic communication signals detected in the space; and

determining utilization of the space during the time period of interest based at least in part on the information indicating the first set of electronic communication signals detected in the space.