WO2009038557A1

WO2009038557A1 - Model-based egress support system

Info

Publication number: WO2009038557A1
Application number: PCT/US2007/020279
Authority: WO
Inventors: Andrzej Banaszuk; Sergey Shishkin; Satish Narayanan; Robert N. Tomastik; Robert E. Labarre; Nathan S. Hariharan; Phillipe Detriche
Original assignee: United Technologies Corporation
Priority date: 2007-09-19
Filing date: 2007-09-19
Publication date: 2009-03-26

Abstract

A model-based egress support system (10) employs models to generate an optimal egress route for occupants in a region. The system includes an egress controller (16) that is connected to receive detection data. The egress controller (16) executes an optimization algorithm (22) to select an egress route based on model-based egress estimates generated by an egress prediction model (18) in response to the detection data.

Description

MODEL-BASED EGRESS SUPPORT SYSTEM

BACKGROUND

The present invention is related to an egress support system, and in particular to a model-based egress controller for supporting egress operations. First responders and other emergency personnel are required to make decisions regarding the allocation of resources when responding to an emergency situation, including how to coordinate resources to save occupants within the building, minimize risk to the first responders, and protect property. In addition, these decisions must typically be made on a time scale consistent with the evolution of the threat, which can require decision-making within seconds or minutes. First responders often have limited information regarding the location of a threat and/or the location of occupants within a building or region. Without further information, first responders may not be able to allocate resources in the most effective manner.

In addition, occupants within a building are typically provided with static instructions regarding egress procedures for exiting a building. For example, egress instructions may be limited to signs within a building illustrating the location of exits and/or exit signs illustrating the presence of an exit. Thus, the egress instructions provided to occupants are based solely on the nearest exit, and do not incorporate dynamic data such as the location of a detected threat. SUMMARY

In one aspect, the present disclosure describes an egress support system that includes an input operably connected to receive detection data. An egress controller includes an optimization algorithm and an egress prediction model. The egress controller executes the optimization algorithm to select an optimal egress route based on the model-based estimates generated by the prediction model in response to the detection data.

In another aspect, the present disclosure describes a method of providing egress support to a region. The method includes acquiring detection data from one or more detection devices. The method further includes generating model-based estimates based on a prediction model and the detection data. The method further includes selecting an optimal egress route for occupants within the region based on the model-based estimates. The method further includes controlling one or more egress instructional devices based on the selected egress route to instruct the movement of occupants within the region. In another aspect, the present disclosure describes an egress controller that includes means for receiving occupant location data from one or more occupant detection devices and means for receiving threat location data from one or more threat detection devices. Based on these inputs and an egress model, the egress controller includes means for generating model-based egress estimates based on an egress prediction model, the occupant location data, and an egress route. The egress control also includes means for selecting an optimal egress route for occupants within the region based on the model-based egress estimates generated with respect to a plurality of egress routes, the occupant location data, and the threat location data. The egress controller further includes means for controlling instructions devices based on the selected egress route to instruct the movement of occupants within the region.

In another aspect, the present disclosure describes a computer readable storage medium encoded with a machine-readable computer program code for selecting an optimal egress route for a region. The computer readable storage medium includes instructions for causing a controller to implement a method that includes acquiring detection data from one or more detection devices. The method further includes generating model-based estimates based on a prediction model and selecting an optimal egress route for occupants within the region based on the model-based estimates.

BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a block diagram of a model-based egress controller.

FIG. 2 is a block diagram of another embodiment of the model-based egress controller.

FIG. 3 is a block diagram of a computer system for implementing the model-based egress controller. DETAILED DESCRIPTION

The present disclosure describes a model-based egress controller that optimizes egress operations for a building or region, hi an exemplary embodiment, an egress controller receives input from one or more sensors describing the location of occupants within the region, location of a detected threat within the region, and/or type of threat detected. The egress controller generates an optimized egress route based on a predictive model and an egress optimization algorithm. hi an exemplary embodiment, the predictive model is a egress prediction model that generates model-based egress estimates regarding the likely propagation of occupants towards selected exits during egress. In particular, the egress prediction model can be used to generate model-based estimates of the time required for all occupants to evacuate. The egress optimization algorithm uses the egress prediction model to model, in real-time, various egress scenarios in which occupants are directed toward a variety of exits. In this way, the egress optimization algorithm is able to determine, based on the modeled scenarios, the egress route that will minimize the time required to evacuate all occupants (i.e. an optimal egress route).

In another exemplary embodiment, the prediction model is a threat prediction model that generates model-based threat prediction estimates regarding the likely propagation of a threat. The egress optimization algorithm uses the threat prediction model to model, in realtime, those egress routes that will be available to occupants based on the model-based threat prediction estimates, hi this way, the egress optimization algorithm is able to determine the egress route that will minimize occupant exposure to a detected threat (i.e., an optimal egress route). Based on the optimized egress route determined by the egress optimization algorithm, the egress controller communicates instructions to occupants within the building or region to direct them along the optimized egress route. In addition, the optimized egress route, along with occupant data and threat detection data, may be provided to first responders to direct their efforts in locating the source of the threat and to aid in the search and rescue of occupants. Finally, the egress controller may provide input to building control operations such as the heating, venting and air-conditioning (HVAC) system and elevator control system to minimize the spread of a detected threat and to aid in the evacuation of occupants.

In addition, the term 'egress route' is used throughout to describe paths available to occupants during egress. The term Optimal egress route' refers to the selection of one or more of the possible egress routes based on some criteria. However, it should be noted that the term 'egress route' and Optimal egress route', although expressed in singular form, may refer to a plurality of individual egress routes. That is, an optimal egress route may include a plurality of individual egress routes designed for occupants located throughout a region.

FIG. 1 is a block diagram of an exemplary embodiment of egress support system 10, which includes occupant detection device(s) 12, threat detection device(s) 14, egress controller 16, egress prediction model 18, threat prediction model 20, egress optimization algorithm 22, egress instructional devices 24, first response devices 26, and building controller 28. Egress controller 16 is connected to receive occupant location data from one or more occupant detection devices 12 and threat detection data from one or more threat detection devices 14. Occupant detection device 12 refers to sensors capable of detecting the location of occupants throughout a region. This may include a binary representation indicating the detected presence of occupants within a particular room or zone of the region (e.g., the room is occupied or the room is un-occupied), or may include additional information concerning the number of occupants detected in a particular room or zone. A variety of devices may be utilized to detect occupants within the region, including motion detection sensors, video detectors, passive infrared sensors, access control devices, elevator load measurements, IT-related techniques (e.g., keystroke detection), as well as other related sensor devices. In addition, many occupants carry active devices, such as active or passive radio frequency identification (RFID) cards, cell phones, or other devices that can be detected to provide data indicative of the occupant's location. Occupant detection device 12 provides occupant location data to egress prediction model 18, which uses the occupant location data as a starting point in generating model-based estimates of the time required to evacuate all occupants.

Threat detection device 14 refers to devices capable of detecting the presence of threats, such as smoke, toxins, gas, or other dangerous or harmful conditions. Once again, the data provided by threat detection device 14 may include a binary representation indicating the detected presence of a harmful condition in a particular location (e.g., harmful condition detected in a room or harmful condition is not detected in a room). In other embodiments, the data provided by threat detection device 14 may include more detailed information regarding the type of harmful condition sensed or concentration of detected condition (i.e., concentration of smoke) at a particular location. Threat detection device 14 may include typical threat detection devices such as smoke alarms or carbon monoxide alarms, or may include non-traditional devices for threat detection such as video devices. Threat detection devices provide threat location data to threat prediction model 20, which uses the threat location data as a starting in generating model-based estimates of the predicted propagation of the detected threat.

In an exemplary embodiment, some detection devices, such as video detection devices, may be used to provide occupant location data as well as threat detection data. Egress prediction model 18 is a mathematical, computer simulation, or statistical model used to predict expected traffic patterns of occupants during an egress condition (i.e., evacuation of the region). Egress prediction model 18 is described in more detail in co- pending PCT application Serial No. filed on even date herewith and entitled "System and Method for Occupancy Estimations" by Robert Tomastik. hi particular, given an initial condition that defines the current location of occupants within the region, and the exits those occupants will be instructed to use, egress prediction model generates model- based estimates of how occupants will move, including an estimate of the time required for all occupants to exit the region. Threat prediction model 20 is a mathematical, computer simulation, or statistical model used to predict the expected propagation of threats through a region. Threat prediction model 20 is described in more detail in co-pending PCT application Serial No. filed on even date herewith and entitled "System and Method for Threat

Propagation Estimation" by Nathan Hariharan. In particular, given an initial condition that defines the current location of the detected threat, threat prediction model 20 generates model-based estimates of the expected propagation of the threat through the region. Therefore, threat prediction model 20 can be used to model the likely path of the threat, including the time it will take a threat to reach various location within the region.

Optimization algorithm 22 employs the predictive capabilities of egress prediction model 18 and threat prediction model 20 to calculate an optimal egress route for occupants within the region. In an exemplary embodiment, the optimal egress route preferably minimizes the time required to evacuate all occupants from the building while also preferably minimizing occupant exposure to the detected threat. The optimal egress route can be used in a variety of ways to aid in the evacuation of a region. For instance, based on a calculated optimized egress route, egress controller 16 may generate egress control instructions that are communicated to occupants within the region via egress instructional devices 24, which guide occupants along the optimal egress route. Egress controller 16 may also provide data (such as the calculated optimal egress route, occupant location data and threat detection data) to first responder devices 26, providing first responders with valuable information regarding the location of a threat, location of occupants, and instructed exit routes communicated to occupants. In addition, egress controller 16 may communicate with building controller 28 to control operations that range from elevator control to heating, venting, and air-conditioning (HVAC) operations of the region. In an exemplary embodiment, optimization algorithm 22 seeks to calculate an egress route that preferably minimizes the time required to evacuate all occupants from a region, subject to the constraint that the egress route selected should minimize occupant exposure to the detected threat. In an exemplary embodiment, optimization algorithm 22 may be formulated in the form of a mixed integer programming problem. In an exemplary embodiment, a linear mixed integer programming algorithm is used to find an egress route that minimizes the total time required to evacuate all occupants. In addition, the linear mixed integer programming algorithm operates within one or more constraints, such as a constraint requiring that occupant exposure to a detected threat should by minimized. In an exemplary embodiment, a constraint related to occupant exposure to a detected threat prevents the selection of an egress route that will expose occupants to the threat. In other embodiments, the exposure to a detected threat may. depend on the type of threat detected. For instance, egress routes directing occupants through smoke-filled hallways may be acceptable in some instances, whereas egress routes directing occupants through areas breached by flames may be prohibited.

Thus, optimization algorithm 22 communicates with egress prediction model 18 and/or threat prediction model 20 to calculate an optimal egress route. In an exemplary embodiment, optimization algorithm 22 instructs egress prediction model 18 to generate model-based estimates, based on a variety of possible egress scenarios. For instance, optimization algorithm 22 may instruct egress prediction model 18 to model the effects of instructing all occupants to move toward a single exit, and another model in which some occupants are instructed to exit through a first exit, and some are instructed to exit through a second exit. Occupant detection data provided by occupant detection device 12 is used to initialize egress prediction model 18, and optimization algorithm 22 instructs egress prediction model 18 to model egress scenarios in which occupants egress through a variety of available exits. The scenario that results in all occupants being evacuated in the shortest amount of time, within the given constraint that occupant exposure to a detected threat should be minimized, is selected by optimization algorithm 22. An optimal egress route selected by optimization algorithm 22 may be dynamically modified based on updated information regarding the correct location of occupants or the location of a detected threat. Thus, the optimal egress route may be dynamically modified as conditions change.

In an exemplary embodiment, optimization algorithm 22 communicates with threat prediction model 20 (or in combination with threat prediction model 20 and egress prediction model 18) to obtain information regarding the anticipated or predicted propagation of the threat through a region. In an exemplary embodiment, optimization algorithm 22 uses model-based threat propagation estimates provided by threat prediction model 20 to determine which exits or areas may be used by occupants, and which areas have been exposed to a particular threat such that instructing occupants to enter the exposed area will violate one of the constraints of optimization algorithm 22.

In an exemplary embodiment, optimization algorithm 22 generates an optimal egress route based only on model-based estimates generated by threat prediction model 20. In another exemplary embodiment, optimization algorithm 22 generates an optimal egress route based on a combination of models generated by egress prediction model 18 and threat predition model 20. Because threat prediction model 20 is predictive, optimization algorithm 22 can use model-based threat propagation estimates that define the likely propagation of the threat at some future time interval to make decisions regarding which exits and passages are available to occupants. For example, if threat prediction model 20 generates an estimate that indicates a detected threat will reach a stairwell in five minutes, optimization algorithm 20 may generate an optimized egress route that instructs those occupants that a model-based egress estimate (generated by egress prediction model 18) suggests can reach and pass through the stairwell within five minutes to use that route. However, occupants that, according to model-based egress estimates, would require more than five minutes to reach or use the stairwell would be instructed to use an alternate route designed by optimization algorithm 22 to prevent exposing those occupants to the likely path of the detected threat.

In other embodiments, optimization algorithm 22 may generate optimized egress routes without predictive knowledge of the expected propagation of the threat (i.e., without input from threat prediction model 20). Rather, optimization algorithm 22 would rely only on the present location of a threat as detected by threat detection device 14 (if available) and predictive estimates of occupant egress generated by egress prediction model 18. Likewise, in other embodiments optimization algorithm may generate optimized egress routes without predictive knowledge of the expected egress of occupants (i.e., without input from egress prediction model 18). Rather, optimization algorithm 22 would rely only on the present location of occupants as detected by occupant detection device 12 and predictive estimates of threat propagation generated by threat prediction model 20. In an exemplary embodiment, the optimized egress route determined by optimization algorithm 22 is used by egress controller 16 to generate egress control instructions. The egress control instructions are communicated to egress instructional devices that are used to communicate the optimized egress route to occupants located within 5 the building or region. These may include visual or auditory devices for communicating instructions to occupants throughout the building. Depending on the location of the occupants, the instructions provided by the visual or auditory devices may vary. Examples of visual devices controlled by egress controller 16 to communicate an optimal egress route to occupants include signs or lights that can be selectively controlled to indicate the

1.0 direction occupants should travel. Visual and auditory instructions may be used alone or in conjunction with one another to communicate to occupants the optimal egress route as determined by egress controller 16.

In an exemplary embodiment, egress controller 16 may also provide data to first- responders regarding the optimal egress route as determined by optimization algorithm 22.

15 First responder devices 26 may be portable devices carried by first responders that are equipped to communicate with egress controller 16. Communication between first responded device 26 and egress controller 16 may be via a telecommunications network, wireless network, or similar communications network. In one exemplary embodiment, the data provided by egress controller 16 may include a visual layout of the building that

20 illustrates visually the location of occupants and detected threats, as well as the optimized egress route generated by egress controller 16. This information aids in the distribution of resources by first responders into the building. For instance, the type of threat detected allows first responders to respond with the proper equipment. Knowledge regarding the location of the threat allows the first responders to target resources more specifically to

25 contain the threat, as well as to avoid placing themselves in danger, and knowledge regarding the location of occupants and instructed egress route allows first responders to target rescue efforts more specifically to maximize the number of people saved.

In an exemplary embodiment, the optimal egress route as determined by optimization algorithm 22 is provided to first responders for review. This allows first

30 responders the opportunity to influence and modify the egress route based on their experience and expertise.

Egress controller 16 may also provide data to building controller 28, which may include systems such as the heating, venting, and air-conditioning (HVAC) control systems and elevator control systems. For instance, in emergency egress situations, elevators may be automatically disabled as a precautionary measure. Oftentimes, however, elevators in a building are capable of operating for some time after detection of a threat. In an exemplary embodiment, egress controller 16 may instruct the elevator control system to continue to operate the elevators until such time that egress controller 16 detects the threat propagating into the elevator shafts. For example, in an exemplary embodiment, threat detection device 14 may be located in the elevator shafts to detect smoke and/or other agents in the elevator shafts. In another exemplary embodiment, threat propagation estimates generated by threat prediction model 20 may be used to predict when a threat will reach a particular elevator shaft. Based on this estimate, optimization algorithm 22 can design an optimized egress route that makes use of the elevator shaft for a period of time before the threat puts the elevator shaft at risk. In this example, egress controller 16 would provide instructions to building controller 28 (or an elevator controller) to cause the elevators to continue to operate while safe. In this way, the elevator may be used as an additional exit for occupants that decreases the overall time required to evacuate occupants from the building.

In addition, based on the sensed location of occupants within the building and the location of a detected threat within the building, egress controller 10 may control the HVAC system to minimize the propagation of the threat towards occupants. For instance, if a chemical agent is detected in one region of the building, egress controller 10, based on the detected location of the chemical agent, may instruct the HVAC control system through building controller 24 to discontinue circulating air from the region in which the chemical agent was detected. This may also be used to prevent or retard the progress of smoke throughout a building. In an exemplary embodiment, control instructions provided to control the operation of the HVAC system or other systems used to control the propagation of a threat are also provided to threat prediction model 20. In this way, threat prediction model 20 is updated to model the propagation of a threat through the building based on control instructions provided to the HVAC system or similar systems.

FIG. 2 illustrates another exemplary embodiment of egress support system 30, which includes occupant detection device(s) 32, threat detection device(s) 34, occupant estimator 36, occupancy estimation algorithm 38, occupant traffic model 40, threat estimator 42, threat propagation algorithm 44, threat propagation model 46, egress controller 48, egress prediction model 50, threat prediction model 52, optimization algorithm 54, egress instructional devices 56, first responder devices 58, and building controller 60.

In the exemplary embodiment shown in FIG. 2, egress controller 48 is connected to receive occupant data (e.g., the location of occupants within a region) from occupant estimator 36, which generates occupancy estimates based on a combination of sensor data provided by occupant detection device 32 and a model-based occupancy estimate generated by occupant traffic model 40. Occupant estimator is described in more detail in co-pending

PCT application Serial No. filed on even date herewith and entitled "System and

Method for Occupancy Estimation" of Robert Tomastik. A benefit of occupant estimator 36 is the ability to provide occupancy estimates despite the loss of sensor data.

In addition, egress controller 48 is connected to receive threat propagation estimates from threat estimator 42, which generates threat propagation estimates based on a combination of sensor data provided by threat detection device 34 and a model-based threat propagation estimate generated by threat propagation model 46. Threat estimator 42 is described in more detail in co-pending PCT application Serial No. filed on even date herewith and entitled "System and Method for Threat Propagation Estimation" of Nathan Hariharan. Once again, a benefit of threat estimator 42 is the ability to provide threat propagation estimates despite the loss of sensor data.

As described with respect to FIG. 1, egress controller 48 may employ model-based egress estimates (generated by egress prediction model 50) and/or model-based threat prediction estimates generates (generated by threat prediction model 52) in generating an optimal egress route.

In an exemplary embodiment, occupant estimates generated by occupant estimator 36 are used to initialize egress prediction model 50. That is, egress prediction model 50 models a variety of egress scenarios based on the occupant estimates provided by occupant estimator 36. As described with respect to FIG. 1, optimization algorithm 54 instructs egress prediction model 50 to generate, in a real-time, a number of egress scenarios in which occupants are modeled evacuating through different exits. Based on the results, a scenario (i.e., optimal egress route) is selected by optimization algorithm 54. that preferably minimizes the amount of time required to evacuate all occupants (while preferably minimizing occupant exposure to the detected threat). In an exemplary embodiment, occupant estimator 36 generates updated occupancy estimates at a defined time-step (e.g., once per second, once every 30 seconds, etc.). In response, each time an updated occupancy estimate is provided to egress controller 48, optimization algorithm 54 re-mns egress scenarios with egress prediction model 50 based on the updated occupancy estimate, hi this way, an optimized egress route may be dynamically modified based on updated occupancy data. In addition, in an exemplary embodiment egress controller 48 provides the optimized egress route as feedback to occupant estimator 36. Because occupant traffic model 40 generates model-based occupancy estimates based on the expected traffic patterns of occupants in the region, providing occupant traffic model 40 with information regarding the optimized egress route (i.e., the egress route communicated to occupants) improves the ability of occupant traffic model 40 to predict the traffic patterns of occupants. That is, occupant traffic model 40 is modified based on the egress instructions provided to occupants via egress instructional devices 56. In this way, the model-based occupant estimates generated by occupant traffic model 40 are improved.

In an exemplary embodiment, egress controller 56 communicates with threat estimator 42 (either alone or in combination with communications with occupant estimator 36). In much the same way that occupant estimates generated by occupant estimator 36 are used to initialize egress prediction model 50, threat estimator generates threat propagation estimates that are used to initialize threat prediction model 52. Based on the threat propagation estimate, which at the very least provides data regarding the current location of a detected threat, threat prediction model 52 generates model-based threat prediction estimates, in real-time, that estimate when a detected threat will propagate throughout a region. As discussed above with respect to FIG. 1, based on threat prediction estimates optimization algorithm 54 is able to comply with constraints that prevent egress routes from putting occupants in danger. In an exemplary embodiment, egress controller 48 provides to threat estimator 42 instructions provided to building controller 60 to modify the propagation of a detected threat. For instance, egress controller 48 may provide instructions to building controller 60 to affect the operation of building systems such HVAC systems to delay or otherwise affect the propagation of a detected threat (e.g., fans located in an area with a detected threat may be shut off to prevent the fans from propagating the threat to other areas). Providing threat estimator 42 with this information allows for threat propagation model 46 to be dynamically modified to account for changes in how the threat will propagate. As discussed with respect to FIG.l, this information may also be used to dynamically modify threat prediction model 52 to account the changes in how the threat will propagate.

Output provided to occupant-based egress instructional devices 56, first responder devices 58, and building controller 60 includes an optimized egress route as determined by egress optimization algorithm 54. In addition, the output provided by egress controller 30 may also include data regarding real-time and near future estimates of occupancy throughout a building or region, predicted propagation of a detected threat, and the likely origin of the detected threat. This information may be particularly useful to first responders to aid in the allocation of resources and personnel. In particular, the likely origin of a detected threat may aid first responders in the containment of the threat. Likewise, information regarding the predicted propagation of a detected threat as well as near future estimates of occupancy throughout a building or region may aid in the allocation of resources to those occupants located in the predicted path of the threat.

In addition, data regarding real-time and near future estimates of occupancy throughout a building or region, predicted propagation of a detected threat, and the likely origin of the detected threat the data may also be provided to building controller 56 to control operations such as elevator operation and HVAC operation. In particular, near future estimates of threat propagation may be useful in determining the length of time elevators within a building may be safely operated despite the detection of a threat. In addition, the likely origin of a detected threat may be useful in controlling HVAC operations to prevent the threat from propagating throughout the building or region.

FIG. 3 illustrates system 70 for generating an optimized egress route base on occupant location data and threat detection data. System. 70 includes controller 72 and computer readable medium 74. In the embodiment shown in FIG. 3, controller 72 executes the steps or processes for calculating an optimized egress route. Thus, the disclosed invention can be embodied in the form of computer or controller implemented processes and apparatuses for practicing those processes. The present invention can also be embodied in the form of computer program code containing instructions embodied in computer readable medium 74, such as floppy diskettes, CD-ROMs, hard drives, or any other computer-readable storage medium, wherein, when the computer program code is loaded into and executed by controller 72, the computer becomes an apparatus for practicing the invention. The present invention may also be embodied in the form of computer program code as a data signal, for example, whether stored in a storage medium 74, loaded into and/or executed by controller 72, or transmitted over some transmission medium, such as over electrical wiring or cabling, through fiber optics, or via electromagnetic radiation, "wherein, when the computer program code is loaded into and executed by a computer, the computer becomes an apparatus for practicing the invention. When implemented on a general-purpose microprocessor, the computer program code segments configure the microprocessor to create specific logic circuits.

For example, in an embodiment shown in FIG. 3, computer readable storage medium 74 may store program code or instructions describing the egress prediction model, the threat prediction model, and the optimization algorithm. The computer program code is communicated to controller 72, which executes the program code to implement the processes and functions described with respect to the present invention.

Although the present invention has been described with reference to preferred embodiments, workers skilled in the art will recognize that changes may be made in form and detail without departing from the spirit and scope of the invention. For example, the present invention has been described with respect to egress operations generated in response to a threat or evacuation mode. In other embodiments the present invention may be used in conjunction with other egress operations. For example, following a concert or sporting event, the present invention may be used to generate optimal egress routes. For example, although a computer system including a processor and memory was described for implementing the egress controller, any number of suitable combinations of hardware and software may be employed for executing the optimization algorithm employed by the egress controller and for storing data and models employed by the optimization algorithm. In addition, the computer system may or may not be used to provide data processing of received sensor data. In some embodiments, the sensor data may be pre-processed before being provided as an input to the computer system responsible for executing the egress controller functions. In other embodiments, the computer system may include suitable data processing techniques to process that provided sensor data (e.g., video recognition software for interpreting and analyzing video data provided by a video detection device).

Furthermore, through the specification and claims, the use of the term 'a' should not be interpreted to mean "only one", but rather should be interpreted broadly as meaning "one or more". The use of sequentially numbered steps used throughout the disclosure does not imply an order in which the steps must be performed. The use of the term "or" should be interpreted as being inclusive unless otherwise stated.

Claims

CLAIMS:

1. An egress support system comprising: an input operably connected to receive occupant detection data, threat detection data, or a combination thereof; an egress controller operably connected to the input and including an optimization algorithm and a prediction model, wherein the egress controller executes the optimization algorithm to select an optimal egress route based on model- based estimates generated by the prediction model in response to the detection data; and an output operably connected to communicate the optimal egress route selected by the optimization algorithm.

2. The egress support system of claim 1, wherein the prediction model is an egress prediction model that generates model-based egress estimates predicting occupant egress based on the occupant location data and a plurality of potential egress routes.

3. The egress support model of claim 2, wherein the optimization algorithm selects the optimal egress route based on the model-based egress estimate that minimizes a time required to evacuate all occupants.

4. The egress support system of claim 1, wherein the prediction model is a threat prediction model that generates a model-based threat estimate that predicts threat propagation based on the threat location data, and wherein the optimization algorithm selects the optimal egress route based on the model-based threat estimate to minimize occupant exposure to the threat.

5. The egress support system of clam. 1, further including: an occupancy estimator operably connected to the input, the occupancy estimator executes an occupancy estimation algorithm that generates an occupancy estimate based on the received occupant location data and a model-based occupancy estimate generated by an occupant traffic model, wherein the occupancy estimate is provided to the egress controller.

6. The egress support system of claim 5, wherein the occupant traffic model is dynamically modified by the egress controller based on the optimal egress route selected by the optimization algorithm.

7. The egress support system of claim 1, further including: a threat propagation estimator pperably connected to the input, the threat propagation estimator executes a threat propagation algorithm that generates a threat propagation estimate based on the received threat location data and a model-based threat propagation estimate generated by a threat propagation model, wherein the threat propagation estimate is provided to the egress controller.

8. The system of claim 1 , wherein the output is operably connected to provide control instructions generated by the egress controller based on the selected egress route to occupant-based egress instructional devices that direct occupants along the selected egress route.

9. The system of claim 1, wherein the occupant-based egress instructional devices include visual signs, auditory devices, or a combination of both.

10. The system of claim 1, wherein the output is operably connected to provide the optimal egress route, the occupant location data, and the threat location data to a first responder device.

11. The system of claim 1, wherein the output is operably connected to communicate building controller instructions generated by the egress controller in response to the threat detection data and the optimal egress route to a building controller,

12. The system of claim 11, wherein the building controller controls the operation of elevators located within the region based on the building controller instructions provided by the egress controller.

13. The system of claim 11, wherein the building controller controls the operation of heating, venting, and air-conditioning (HVAC) systems located within the region based on the building controller instructions provided by the egress controller.

14. A method of providing egress support to a region, the method comprising: acquiring detection data from one or more devices, wherein the detection data is occupant detection data, threat detection data, or a combination thereof; generating a model-based estimate based on a predictive model and the detection data; selecting an optimal egress route for occupants within the region based on the model-based estimates; and controlling one or more egress instructional devices based on the selected optimal egress route to instruct the movement of occupants within the region.

15. The method of claim 14, wherein generating the model-based estimate includes generating model-based egress estimates based on an egress prediction model, a plurality of possible egress routes, and the occupant detection data, wherein each model-based egress estimate includes an estimated time to evacuate occupants based on one of the plurality of possible egress scenarios, and wherein selecting an optimal egress route includes selecting one of the plurality of possible egress routes based on the model-based egress estimate that minimizes the estimated time to evacuate the occupants.

16. The method of claim 14, wherein generating the model-based estimate include generating model-based threat prediction estimates based on a threat prediction model and the threat detection data, and wherein selecting an optimal egress route includes selecting an egress route that minimizes occupant exposure to a detected threat based on the model- based threat prediction estimate.

17. The method of claim 15, wherein generating the model-based estimated further includes generating a model-based threat prediction estimate based on a threat prediction model, and wherein selecting an optimal egress route includes selecting an egress route that minimizes the estimated time to evacuate the occupants and minimizes occupant exposure to a detected threat based on the model-based threat prediction estimate.

18. The method of claim 14, wherein controlling one or more egress instructional devices includes controlling auditory devices and visual devices to communicate the selected egress route to occupants within the region.

19. An egress controller comprising: means for acquiring occupant location data from one or more occupant detection devices; means for acquiring threat location data from one or more threat detection devices; means for generating model-based egress estimates based on an egress prediction model, the occupant location data, and an egress route; means for selecting an optimal egress route for occupants within the region based on the model-based egress estimates generated with respect to a plurality of egress routes, the occupant location data, and the threat location data; and means for providing based on the selected egress route, instructions for movement of occupants within the region.

20. The egress controller of claim 19, wherein the means for selecting an optimal egress route is an optimization algorithm that causes the means for generating model-based egress estimates to generated a plurality of model-based egress estimates based on egress routes provided by the optimization algorithm.

21. The egress controller of claim 20, wherein the model-based egress estimates include time data describing with respect to each model-based egress estimate the time required for all occupants to evacuate a region; and the means for selecting an optimal egress route selects an egress route based on the time data to minimize the time required to evacuate all occupants.

22. The egress controller of claim 21, wherein the means for selecting the optimal egress route selects an egress route based on the threat location data to minimize occupant exposure to a detected threat.

23. The egress controller of claim 19, wherein the means for acquiring occupant location data includes: an occupancy estimator that generates an occupant estimate based on the occupant location data and a model-based occupant estimate generated by an occupant estimation model, wherein the model-based egress estimates are generated based on the occupant estimate.

24. The egress controller of claim 19, wherein the means for acquiring threat location data includes: a threat propagation estimator that generates a threat propagation estimate based on the threat location data and a model-based threat propagation estimate generated by a threat propagation mode, wherein the optimal egress route is selected based on the threat propagation estimate.

25. A computer readable storage medium encoded with a machine-readable computer program code for selecting an optimal egress route for a region, the computer readable storage medium including instructions for causing a controller to implement a method comprising: acquiring detection data from one or more detection devices, wherein the detection data is occupant detection data, threat detection data, or a combination thereof; ^• generating model-based estimates based on a prediction model and the detection data; and selecting an optimal egress route for occupants within the region based on the model-based estimates.