US20210398063A1

US20210398063A1 - Method and system for accurately estimating amount of materials in stores

Info

Publication number: US20210398063A1
Application number: US17/350,948
Authority: US
Inventors: Theodore Trebaol; Louis Trebaol
Original assignee: Individual
Current assignee: Individual
Priority date: 2020-06-18
Filing date: 2021-06-17
Publication date: 2021-12-23
Also published as: WO2021258080A1; WO2021257957A1; US20210398062A1

Abstract

A method and system are disclosed for estimating quantities of stored materials. Stores data can be previously collected by a person (manual stores) using manually collected data and higher fidelity stores data can be collected by automated sensors from other similar stores (monitored stores). The system takes as initial input historical data from a set of monitored stores and historical data for a set of manual stores. The system continues to receive monitored stores data and manual stores data over time. The system applies a method to the input data to generate estimates of remaining material quantities in the set of manual stores since their latest manual data collection.

Description

CLAIM OF BENEFIT TO PRIOR APPLICATION

This application claims benefit to U.S. Provisional Patent Application 63/040,759, entitled “METHOD AND SYSTEM FOR ACCURATELY ESTIMATING AMOUNT OF MATERIALS IN STORES,” filed on Jun. 18, 2020. The U.S. Provisional Patent Application 63/040,759 is incorporated herein by reference.

BACKGROUND

Material supply chains connect nodes of producers, distributors, vendors, and consumers, and include material storage assets at each node. Material producers and distributors may enter into arrangements that guarantee a consumer's access to the materials by allowing the producers and distributors to control the schedule of interactions with the store (e.g. resupplying a desired quantity of the material in the store). In this situation, producers and distributors can choose to install remote sensors on stores that measure and report material quantities so that interactions with the store may be scheduled to adequately manage the store (e.g. ensuring the store does not run out of material). Directly measuring a store can result in the most efficient supply chain. User credentials from logins can be copied by others. When such login credentials are copied, they can also be distributed to others who may then access secure data. Also the data is usually never encrypted and thereby has the means of being stolen without having to do much work in deciphering what is in the file.

BRIEF DESCRIPTION OF THE DRAWINGS

However, not all stores may be measured using sensors. Some stores need to be manually managed by the producers and distributors when remote sensing is not an option. Furthermore, some consumers prefer to manage their own stores by manually requesting some or all interactions (e.g. storage asset maintenance or resupplying material). Producers and distributors may attempt to fit these requests into their service schedules as soon as possible because any delays may negatively impact the consumer's access to materials. Attempting to anticipate these consumer requests can lead to operating supply chains with unnecessarily excessive capacity.

Features and advantages of embodiments of the claimed subject matter will become apparent as the following Detailed Description proceeds, and upon reference to the Drawings, in which:

FIG. 1 illustrates a multi-store monitoring environment, m accordance with some embodiments of the present disclosure.

FIG. 2 illustrates a flowchart of an example store monitoring method, in accordance with some embodiments of the present disclosure.

Although the following Detailed Description will proceed with reference being made to illustrative embodiments, many alternatives, modifications, and variations thereof will be apparent in light of this disclosure.

DETAILED DESCRIPTION

As noted above, there are some non-trivial issues with monitoring multiple stores that include stores without automatic monitoring capability (e.g., via use of a sensor). As used herein, the term “monitored store” refers to a store that is monitored using a device (such as a sensor) that can automatically provide data about the state of the store. The term “unmonitored store” or “manual store” refers to a store that is not automatically monitored, and thus data collection regarding the state of the store is performed manually (e.g., manual inspection of the store or data received from the consumer about the state of the store).
In some cases, producers and/or distributors of a store's material may manually collect data from the unmonitored stores via personal inspection or requests to the consumer for the data. Material consumption models can be designed based on the manually collected data to attempt to estimate future or current inventory levels in the store. Such models can include constant consumption models, seasonal consumption models, or environmentally driven consumption models (e.g. attempting to estimate heating fuel consumption from a tank using weather information measured in the vicinity of the consumer). However, such current estimation techniques can result in supply chains serving manual stores that are either less efficient than those serving monitored stores or materials being depleted before they can be resupplied.
Accordingly, systems and methods are disclosed herein for providing more accurate models of the inventory levels of unmonitored stores. According to some embodiments, methods described herein combine high-fidelity monitored stores data (e.g. from sensor readings) with manual stores data and generates accurate high-fidelity level estimates of manual stores to maximize the efficiency of managing all of the stores (e.g., monitored and unmonitored). The embodiments described herein provide an alternative solution to having to install a sensor on each store and allow material producers and/or distributors to install remote sensors on a subset of their total stores population and while virtually monitoring other manual stores to provide current estimates of the manual stores' levels with the same fidelity as a physical sensor.
Some embodiments of the present disclosure provide a method and system for estimating inventory levels in infrequently monitored stores using data collected when interacting with these stores and data collected from other similar stores which are monitored with automated sensors. The provided inventory level estimates increase the amount of insight producers and distributors of materials have into their consumer markets. An example of clients that might apply the disclosed systems or methods include industrial chemical producers and distributors storing liquid or gas materials in tanks. In such an industry, materials are transported by vehicles from tanks at producer and distributor locations to tanks at consumer locations. A producer is more cost effective if they can avoid storing excess material between production and distribution. A distributor is more cost effective if they minimize the number of resources used to transport materials to consumer storage tanks. Accurate inventory level estimates enable such a distributor to use fewer vehicles and technicians, and for less time, to meet the same market demand. In some embodiments, different models can be created for different time periods, such as different seasons. For example, a first set of models may be used for particular stores during winter months while a second set of models may be used for the stores during the summer months. The different models can better reflect the different usage rates based on the weather and may be especially useful for stores that hold heating fuel, such as propane.
In an example embodiment, a method for estimating inventory level of an unmonitored store includes receiving first data from one or more sensors providing inventory levels of one or more monitored stores; receiving second data, which may comprise at least one of historical resupply dates and historical inventory level measurements of the unmonitored store; generating a model to estimate the inventory level of the unmonitored store based at least on the first data and the second data; and estimating a current inventory level for the unmonitored store based on the model.
In another example embodiment, a method for estimating a time at which an unmonitored store's inventory will be depleted includes receiving first data from one or more sensors providing inventory levels of one or more monitored stores; receiving second data associated with the unmonitored store, where the second data may include at least one of historical resupply dates and historical inventory level measurements of the unmonitored store; generating a model to estimate the time at which the unmonitored store's inventory will be depleted based at least on the first data and the second data; and estimating the time at which the unmonitored store's inventory will be depleted based on the model.
The methods described herein may be applied to situations beyond fluid monitoring. According to some embodiments, the monitoring methods disclosed herein may be applied to any measurable inventory, such as, for example, boxes on grocery store shelves or merchandise at any store. The techniques may be embodied in devices, systems, methods, or machine-readable mediums, as will be appreciated.
The description uses the phrases “in an embodiment” or “in embodiments,” which may each refer to one or more of the same or different embodiments. Furthermore, the terms “comprising,” “including,” “having,” and the like, as used with respect to embodiments of the present disclosure, are synonymous. When used to describe a range of dimensions, the phrase “between X and Y” represents a range that includes X and Y.
Spatially relative terms, such as “beneath,” “below,” “lower,” “above,” “upper,” and the like, may be used herein for ease of description to describe one element or feature's relationship to another element (s) or feature (s) as illustrated in the figures. The spatially relative terms are intended to encompass different orientations of the device in use or operation in addition to the orientation depicted in the figures. The apparatus may be otherwise oriented (rotated 90 degrees or at other orientations) and the spatially relative descriptors used herein may likewise be interpreted accordingly.
FIG. 1 illustrates an example monitoring environment 100. Monitoring environment includes a plurality of stores 102 made up of unmonitored stores 104 and monitored stores 106. Each of the stores in plurality of stores 102 can include any quantifiable material such as any liquid any gas, or any objects (e.g., consumer products). Some example materials to be held within the stores include liquid nitrogen, oxygen, propane, natural gas, oil, etc. The stores themselves can be any size and do not all have to be the same size.
Each of monitored stores 106 includes a sensor 108 that is used to collect information about the associated store, according to an embodiment. Sensor 108 may be an electronic sensor that measures an inventory level within the store and transmits this inventory level across network 110 to a computing device 111 designed to receive measurements from each of sensor 108. In some examples, sensor 108 can measure other parameters such as temperature, humidity, pressure, or GPS location and can send this data as well across network 110.
According to an embodiment, computing device 111 collects information across network 110 from different sources and executes software to estimate inventory levels and/or times at which inventory will be depleted for one or more of unmonitored stores 104. Network 110 may represent the Internet or any other network infrastructure. Some examples of network 110 include cellular communication, satellite communication, WIFI communication, radiofrequency (RF) communication, and fiber-optic communication.
Computing device 111 can be any computer system, such as a workstation, desktop computer, server, laptop, handheld computer, tablet computer (e.g., the iPad® tablet computer), mobile computing or communication device (e.g., the iPhone® mobile communication device, the Android™ mobile communication device, and the like), VR device or VR component (e.g., headset, hand glove, camera, treadmill, etc.) or other form of computing or telecommunications device that is capable of communication and that has sufficient processor power and memory capacity to perform the operations described in this disclosure. A distributed computational system can be provided including a plurality of such computing devices. According to some embodiments, computing device 111 includes a processor 112, memory 114, network I/O 116, and a I/O device 118. Computing device 111 may also have a graphical user interface (GUI) 120 that may include a display and a user input device. In some embodiments, GUI 120 represents a command-line interface.
According to some embodiments, memory 114 represents any transitory or non-transitory internal memory or computer-readable media having encoded thereon one or more computer-executable instructions or software for implementing techniques as variously described in this disclosure. Memory 114 can include a computer system memory or random access memory, such as a durable disk storage (which can include any suitable optical or magnetic durable storage device, e.g., RAM, ROM, Flash, USB drive, or other semiconductor-based storage medium), a hard-drive, CD-ROM, or other computer readable media, for storing data and computer-readable instructions or software that implement various embodiments as taught in this disclosure. Memory 114 can be provided on the computing device 111 or provided separately or remotely from the computing device 111.
According to some embodiments, processor 112 of the computing device 111 is configured to execute commands to estimate inventory levels and/or times at which inventory will be depleted for one or more unmonitored stores within monitoring environment 100. Processor 112 may be used to execute computer-readable and computer-executable instructions or software stored in memory 114 as well as other programs for controlling system hardware. Network I/O 116 can be any appropriate network chip or chipset which allows for wired or wireless connection between computing device 111 and communication network 110 in order to send/receive data from other computing devices and resources.
A user can interact with computing device 111 through I/O device 118, such as a screen or monitor, which can display one or more user interfaces or images, including data regarding fill levels or resupply schedules for unmonitored stores, as provided in accordance with some embodiments. Computing device 111 can include I/O devices 118 for receiving input from a user, for example, a keyboard, a joystick, a game controller, a pointing device (e.g., a mouse, a user's finger interfacing directly with a touch-sensitive display device, etc.), or any suitable user interface, including an AR headset. The computing device 111 may include any other suitable conventional I/O peripherals. In some embodiments, computing device 111 includes or is operatively coupled to various suitable devices for performing one or more of the aspects as variously described in this disclosure.
According to some embodiments, computing device 111 receives measurement data from sensors 108 of monitored stores 106. The data may include current inventory levels of monitored stores 106 and can be received at a fairly high frequency depending on the configurable nature of sensor 108. For example, sensor 108 may be designed to transmit data multiple times a day or once a day. In some embodiments, sensor 108 takes multiple data measurements at a first frequency but transmits the data at a second frequency lower than the first frequency. The frequency of data transmission may depend on what material is being held within a given store and how often it is being used.
Computing device 111 also receives historical data from other sources regarding both monitored stores 106 and unmonitored stores 104, according to an embodiment. For example, computing device 111 receives historical manual stores data 122, which may include past dates that one or more of unmonitored stores 104 was resupplied and/or how much was resupplied. Historical manual stores data 122 may also include manual fill-level readings that had been taken of one or more of the unmonitored stores 104. Computing device may also receive historical monitored stores data 124, which may include past dates that one or more of the monitored stores 106 was resupplied and/or how much was resupplied. Historical monitored stores data 124 may also include any past sensor readings from one or more of the monitored stores 106.
According to some embodiments, computing device 111 also receives most recent data with regards to the monitored stores 106 and unmonitored stores 104 in order to continually update and refine models for estimating the inventory levels and/or times at which inventory will be depleted in the monitored stores 106. For example, computing device 111 may receive latest measurements from any of sensors 108 to provide the most recent inventory level measurements of any of monitored stores 106. Computing device 111 may also receive new manual stores data 126 corresponding to a latest manual measurement taken from one or more of unmonitored stores 104.
According to some embodiments, each of historical manual stores data 122, historical monitored stores data 124, and new manual stores data 126 is received from a second computing device, such as a computing device associated with the plurality of stores 102. The second computing device may be owned by a consumer that uses the materials stored in plurality of stores 106, and thus the consumer provides the additional data about their stores to computing device 111. In some embodiments, the additional data is provided to computing device 111 from different sources. For example, a consumer may provide some of the data (e.g., new manual stores data 126) via their smartphone or other computer, while a separate computer keeps track of the historical manual stores data 122 and historical monitored stores data 124 and sends this data automatically to computing device 111.
FIG. 2 is a flow diagram of an example method 200 for estimating the inventory levels of one or more manual stores, in accordance with an embodiment. Method 200 may be performed, for example, in whole or in part by computing device 111 as described with respect to FIG. 1. The operations, functions, or actions described in the respective blocks of example method 200 may also be stored as computer-executable instructions in a non-transitory computer-readable medium, such as a memory and/or a data storage of a computing system. As will be further appreciated in light of this disclosure, for this and other processes and methods disclosed herein, the functions performed in method 200 may be implemented in a differing order. Additionally, or alternatively, two or more operations may be performed at the same time or otherwise in an overlapping contemporaneous fashion.
Method 200 includes the collection of both manual data at block 202 (from one or more unmonitored stores) and monitored data at block 204 (from one or more monitored stores), according to an embodiment. The manual data may include historical data collected from N manual stores, where the historical data includes one or more of dates that the manual store was resupplied, the final liquid level amount of the store after being resupplied, and any liquid level measurements taken between resupplies. Similarly, the monitored data may include historical data collected from M monitored stores over the same time period, where the historical data includes one or more of dates that the monitored store was resupplied, the final liquid level amount of the store after being resupplied, and any liquid level measurements taken between resupplies. New data may be added to the collections of manual and monitored stores data as it is sent from a client.
According to an embodiment, the collected historical datasets are converted into a series of time-stamped inventory level measurements for each store (both manual stores at block 206 and monitored stores at block 208). For the example application of an industrial chemical distributor using this system, the manually collected data could be taken when tanks are resupplied, and data collected from monitored tanks it taken automatically by a remote sensor. Each measurement for each tank may also be marked as to which tank it corresponds.
According to an embodiment, U measurements may be collected for each manual store. Then, for each manual store, method 200 proceeds to block 210 where the amount of materials consumed between each measurement is calculated by taking the difference between the inventory levels (e.g. gallons of chemicals) and saving this difference as a consumption series. For a total of N manual stores, the consumption series each have a length of (U−1). The amount of time between each manual store's measurements may be calculated and saved as a (U−1) series of time periods with the consumption series, according to an embodiment. Similarly, 5 measurements may be collected for each monitored store. Then, for each monitored store, method 200 proceeds to block 212 where the amount of materials consumed between each sensor measurement is calculated by first removing changes in inventory level that occur when the materials are replenished and then by taking the difference between the inventory levels (e.g. gallons of chemicals), according to an embodiment. Each of these may be saved for a total of M consumption series with length (5−1).
According to an embodiment, method 200 continues to block 214 where, for each manual store, the monitored store's consumption series data is resampled by calculating the cumulative sum of monitored material consumed between each manual store's level measurement, resulting in M re-sampled monitored stores consumption series now each with length (U−1). In the event a monitored store did not provide data during the time period between any two manual store measurements, the value for re-sampled consumption series for that time period may be set to a null value. These M series are combined with their corresponding manual store's consumption series to generate N matrices of size (U−1) by (M+1), according to an embodiment.
According to an embodiment, method 200 continues to block 216 where a threshold T1 is applied to the number of non-null values in each manual store's matrix of consumption series. The threshold may be applied to minimize the probability of spurious correlations in later operations. For each manual store's consumption matrix, any monitored store's re-sampled consumption series with fewer than T1 non-null values may be removed from the matrix.
According to an embodiment, method 200 continues to block 218 where, for each manual store's consumption matrix with M1 remaining re-sampled monitored consumption series, the manual store's consumption series are paired with each of the M1 monitored consumption series and both M1 correlation coefficients and null-hypothesis probabilities are determined.
According to an embodiment, at block 220 thresholds T2 and T3 are applied to the correlation coefficients and null-hypothesis probabilities. This may be performed to minimize the computational complexity of subsequent operations. For each manual store's consumption matrix, any correlation coefficient below T2 or null-hypothesis probability above T3 causes the monitor consumption series used to generate them to be removed from the matrix.
Method 200 proceeds to block 222 where a regression analysis is used to model the manual store's consumption series as a continuous function of the monitored store's re-sampled consumption series for each manual store's consumption matrix and for each M2 remaining re-sampled monitored consumption series, according to an embodiment. This generates M2 equations, which may be referred to as models, each with one independent variable and one dependent variable with the general form y=f(x) where y is the estimated amount of material consumed by the manual store and x is the amount consumed by the monitored store during the same time period.
According to an embodiment, method 200 proceeds to block 224 where, for each manual store's consumption matrix and for each M2 model, the re-sampled monitor consumption series is used as input to their respective model and each model's estimate of the manual store's consumption series is calculated. Here each re-sampled monitor consumption series is already of length (U−1), and so each model's output is the same length as the manual store's consumption sense.
According to an embodiment, method 200 proceeds to block 226 where, for each manual store's consumption matrix, for each M2 model, and for each (U−1) consumption estimates (U−1) error values are calculated as follows:
$\frac{\begin{matrix} (Model Estimated Consumption - \\ Actual Manual Store Consumption) \end{matrix}}{Number of Days in Time Period}$
According to an embodiment, for each manual store's consumption matrix and for each M2 model, the average of the error values and the variance the error values are calculated at block 228.
According to an embodiment, method 200 continues to block 230 where, for each manual store's consumption matrix and for each M2 model, the models are ranked in order of decreasing error valance.
According to an embodiment, for each manual store's consumption matrix, a threshold M3 is used at block 232 and all but the top M3 models are removed at block 234. The remaining models correspond to those which have the smallest error variance. If there are fewer than M3 models following the operations performed in blocks 230, 232, and 234, then all remaining models are taken.
According to an embodiment, method 200 proceeds to blocks 236 and 238 where each manual store's set of models are combined and saved along with unique identifiers of the monitored stores from which they were derived for comparison with future models that may be built with additional manual stores data.
According to an embodiment, for each manual store, if there are models from previous iterations collected at block 238 that were derived from the associated monitored store then the terms of the new model are compared to the terms of the previous model at block 240. For example, if the previous model took the form of y=A1x3+B1x2+C1x and the new model tool the form y=A2x3+B2 x2+C2 x, then the comparison of terms would include A1 to A2, B1 to B2, and C1 to C2. If any of these term pairs are different by more than a threshold M4, then an alert may be generated so that a user may take a closer look at the data used to generate the model before accepting its level estimate.
According to an embodiment, at block 242 the latest manual store measurement is used to initialize a digital copy of each manual store's inventory.
According to an embodiment, as new monitor data is received over time as illustrated by block 244, the manual store's consumption is estimated using data from the monitored stores in each manual store's set of models at block 246. According to some embodiments, this process may be repeated as often as, but no faster than, monitor data is received to generate level estimates at the similar fidelity as the monitor data. This process may also be repeated at an intermediate frequency (e.g. daily) to cover any reporting time period. For each manual store, for each model, and for each reporting time period the method includes identifying which monitored tanks in the combined model have data available for use within that time period. For each K monitored tank that has data available with standard deviation of error CJ, the system calculates weight terms co as follows:
$\frac{1}{?}$ $? - \sum_{i = 1}^{K} ? \frac{1}{2}$ $? indicates text missing or illegible when filed$
The output of the i'th model is weighted by multiplying its output by its respective weight term coi. The sum of the weighted outputs for each reporting time period are taken together to estimate the manual store's consumption during the reporting time period, according to an embodiment. The estimated consumption may then be subtracted from the digital copy of the manual store to calculate the current level estimate for the manual store at block 248. In some embodiments, this value also updates the digital copy of the manual store. At block 250, the level estimates for the manual store may be saved for future use, outputted for viewing by a user, or used along with the manual historical data to determine a future time at which the inventory will be depleted for the manual store.
Unless specifically stated otherwise, it may be appreciated that terms such as “processing,” “computing,” “calculating,” “determining,” or the like refer to the action and/or process of a computer or computing system, or similar electronic computing device, that manipulates and/or transforms data represented as physical quantities (for example, electronic) within the registers and/or memory units of the computer system into other data similarly represented as physical quantities within the registers, memory units, or other such information storage transmission or displays of the computer system. The embodiments are not limited in this context.
Numerous specific details have been set forth herein to provide a thorough understanding of the embodiments. It will be understood in light of this disclosure, however, that the embodiments may be practiced without these specific details. In other instances, well known operations and components have not been described in detail so as not to obscure the embodiments. It can be appreciated that the specific structural and functional details disclosed herein may be representative and do not necessarily limit the scope of the embodiments. In addition, although the subject matter has been described in language specific to structural features and/or methodological acts, it is to be understood that the subject matter defined in the appended claims is not necessarily limited to the specific features or acts described herein. Rather, the specific features and acts described herein are disclosed as example forms of implementing the claims.

Claims

I claim:

1. A method for estimating inventory level of an unmonitored store, the method comprising:

receiving first data from one or more sensors providing inventory levels of one or more monitored stores;

receiving second data associated with the unmonitored store;

generating a model to estimate the inventory level of the unmonitored store based at least on the first data and the second data; and

estimating a current inventory level for the unmonitored store based on the model.

2. The method of claim 1, wherein the second data comprises at least one of historical resupply dates for the unmonitored store and historical inventory level measurements for the unmonitored store.

3. The method of claim 1 or 2, further comprising measuring the first data at a first frequency and measuring the second data at a second frequency lower than the first frequency.

4. The method of any one of claims 1-3, wherein the model is a first model, and the generating comprises generating the first model for a first seasonal time period, and the method comprises generating a second model for a second seasonal time period different from the first seasonal time period.

5. The method of any one of claims 1-4, further comprising:

receiving third data associated with the unmonitored store; and

updating the model to more accurately predict the inventory level of the unmonitored store based on the third data.

6. The method of claim 5, further comprising receiving fourth data from the one or more sensors associated with the one or more monitored stores.

7. The method of claim 6, further comprising applying the fourth data to the model to estimate a consumption rate of the unmonitored store.

8. The method of claim 6 or 7, wherein receiving the fourth data comprises receiving a most recent inventory level measurement for the one or more monitored stores.

9. A computer program product including one or more non-transitory machine-readable media having instructions encoded thereon that when executed by at least one processor causes a process for estimating inventory levels of an unmonitored store to be carried out, the process comprising:

receiving second data associated with the unmonitored store;

generating a model to predict the inventory level of the unmonitored store based at least on the first data and the second data; and

10. The computer program product of claim 9, wherein the second data comprises at least one of historical resupply dates for the unmonitored store and historical inventory level measurements for the unmonitored store.

11. The computer program product of claim 9 or 10, the process further comprising measuring the first data at a first frequency and measuring the second data at a second frequency lower than the first frequency.

12. The computer program product of any one of claims 9-11, wherein the model is a first model, and the generating comprises generating the first model for a first seasonal time period, and the method comprises generating a second model for a second seasonal time period different from the first seasonal time period.

13. The computer program product of any one of claims 9-12, the process further comprising:

receiving third data associated with the unmonitored store; and

14. The computer program product of claim 13, the process further comprising receiving fourth data from the one or more sensors associated with the one or more monitored stores.

15. The computer program product of claim 14, the process further comprising applying the fourth data to the model to estimate a consumption rate of the unmonitored store.

16. The computer program product of claim 14 or 15, wherein receiving the fourth data comprises receiving a most recent inventory level measurement for the one or more monitored stores.

17. A method for estimating a time at which an unmonitored store's inventory will be depleted, the method comprising:

receiving second data associated with the unmonitored store;

generating a model to estimate the time at which the unmonitored store's inventory will be depleted based at least on the first data and the second data; and

estimating the time at which the unmonitored store's inventory will be depleted based on the model.

18. The method of claim 17, wherein the second data comprises at least one of historical resupply dates for the unmonitored store and historical inventory level measurements for the unmonitored store.

19. The method of claim 17 or 18, further comprising measuring the first data at a first frequency and measuring the second data at a second frequency lower than the first frequency.

20. The method of any one of claims 17-19, wherein the model is a first model, and the generating comprises generating the first model for a first seasonal time period, and the method comprises generating a second model for a second seasonal time period different from the first seasonal time period.

21. The method of any one of claims 17-20, further comprising:

receiving third data associated with the unmonitored store; and

updating the model to more accurately estimate the time at which the unmonitored store's inventory will be depleted based on the third data.

22. The method of claim 21, further comprising receiving fourth data from the one or more sensors associated with the one or more monitored stores.

23. The method of claim 22, further comprising applying the fourth data to the model to estimate a consumption rate of the unmonitored store.

24. The method of claim 22 or 23, wherein receiving the fourth data comprises receiving a most recent inventory level measurement for the one or more monitored stores.

25. A computer program product including one or more non-transitory machine-readable media having instructions encoded thereon that when executed by at least one processor causes a process for estimating a time at which an unmonitored store's inventory will be depleted to be carried out, the process comprising:

receiving second data associated with the unmonitored store;

26. The computer program product of claim 25, wherein the second data comprises at least one of historical resupply dates for the unmonitored store and historical inventory level measurements for the unmonitored store.

27. The computer program product of claim 25 or 26, the process further comprising measuring the first data at a first frequency and measuring the second data at a second frequency lower than the first frequency.

28. The computer program product of any one of claims 25-27, wherein the model is a first model, and the generating comprises generating the first model for a first seasonal time period, and the method comprises generating a second model for a second seasonal time period different from the first seasonal time period.

29. The computer program product of any one of claims 25-28, the process further comprising:

receiving third data associated with the unmonitored store; and

30. The computer program product of claim 29, the process further comprising receiving fourth data from the one or more sensors associated with the one or more monitored stores.

31. The computer program product of claim 30, the process further comprising applying the fourth data to the model to estimate a consumption rate of the unmonitored store.

32. The computer program product of claim 30 or 31, wherein receiving the fourth data comprises receiving a most recent inventory level measurement for the one or more monitored stores.