System and method for estimating the number of people in a smart building

• US 9,618,918 B2
• Filed: 03/07/2016
• Issued: 04/11/2017
• Est. Priority Date: 07/13/2015
• Status: Expired due to Fees

• Current Assignee: O'Keeffe, James Thomas
• Original Assignee: O'Keeffe, James Thomas
• Inventors: O'Keeffe, James Thomas
• Primary Examiner(s): LAU, TUNG S
• Application Number:
  US15/063,142
• Publication Number:
  US 20170017214A1
• Time in Patent Office: 400 Days
• Field of Search: 702189
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 2219/2614 HVAC, heating, ventillation, climate control
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

This disclosure concerns estimating the number of people within building-based device networks. Increasingly, local area network (LAN) devices are capable of sensing people (e.g. smart appliances). Features of building layout, and physical arrangement of devices provide that only a subset of sensor data patterns can be accounted for by a single user (e.g. simultaneous activation of a smart TV and a motion sensor). Over time, a central controller (e.g. LAN gateway) can gather these building-specific sensor patterns and learn single occupant characteristics for a network (e.g. features or boundaries of single occupant capability). In one embodiment, the central controller generates a single occupancy criterion (e.g. a building-specific test), comprising single occupant characteristics, operable to test a subset of sensor data for single occupancy. The central controller divides devices into a minimum number of groups, such that each group satisfies the single occupancy criterion and estimates the number of occupants accordingly.

Claims

1. A computer implemented method for estimating the number of people in a region within an enclosed space comprising:

obtaining a set of sensor data from a plurality of devices within a network of devices, wherein the set of sensor data comprises for each device in the plurality of devices corresponding sensor data indicative of a measure of occupancy within at least a portion of the region,

obtaining a single occupancy criterion comprising a plurality of single occupant characteristics,

wherein the single occupancy criterion functions to identify whether a candidate subset of the set of sensor data can be accounted for by a single occupant,

wherein the single occupancy criterion further comprises a test based on the plurality of single occupancy characteristics, and wherein the single occupancy criterion is satisfied by the candidate subset of the set of sensor data when the candidate subset passes the test,

processing one or more subsets of the set of sensor data according to the single occupancy criterion,

wherein the processing identifies a set of subsets of devices such that for each subset of devices in the set of subsets of devices the corresponding subset of the set of sensor data satisfies the single occupancy criterion, and such that each device in the plurality of devices is included in at least one subset of devices in the set of subsets of devices, and

estimating the number of people in the region within the enclosed space according to the number of subsets of devices in the set of subsets of devices.

2. The method of claim 1:

wherein each of the plurality of single occupant characteristics is a rule operable to be evaluated on the candidate subset of the set of sensor data and which when satisfied by the candidate subset of the set sensor data indicates that the candidate subset of the set of sensor data can be accounted for by a single occupant, and

wherein at least one of the plurality of single occupant characteristics is generated by a computer and based on processing training data from at least one of the plurality of devices by the computer.

3. The method of claim 1 wherein the single occupancy criterion comprises at least one single occupant characteristic that is a concurrent characteristic and at least one single occupant characteristic that is a sequential characteristic.

4. The method of claim 1 wherein at least one single occupant characteristic describes a minimum transition time between at least two devices from the plurality of devices that can be accounted for by a single occupant.

5. The method of claim 1 wherein at least one of the plurality of single occupant characteristics is generated at least in part by a computer coupled to the network of devices.

6. The method of claim 1:

wherein the processing identifies the set of subsets of devices by iteratively performing, the steps of;

a. generating a new subset of devices comprising the largest possible number of devices from the plurality of devices that are not already included in a subset of devices in the set of subsets of devices, and such that the single occupancy criterion is satisfied by the corresponding subset of the set of sensor data for all devices in the new subset of devices,

b. adding the new subset of devices to the set of subsets of devices.

7. The method of claim 1 wherein each of the plurality of single occupant characteristics defines how to generate a corresponding score when evaluated on the candidate subset of the set of sensor data, and

wherein the test defines a scoring criterion for passing the test, the scoring criterion using the score corresponding to at least one of the plurality of single occupancy characteristics.

8. The method of claim 1 wherein for each of the plurality of devices the corresponding sensor data is obtained in short range wireless signals having a range of less than 800 meters.

9. The method of claim 1 wherein the network of devices is a local area network and wherein the processing to identify the set of subsets of devices is performed by a computer that is part of the local area network.

10. The method of claim 1 further comprising:

estimating the number of people in the region according to the number of subsets of devices in the set of subsets of devices and according to the number of subsets of devices in the set of subsets of devices at an earlier time.

11. The method of claim 1 wherein the test in the single occupancy criterion is based on at least four single occupant characteristics.

12. The method of claim 1 further comprising:

prior to generating the set of subsets of devices,

varying a detection threshold to generate a varied detection threshold, wherein the varied detection threshold is based at least in part on a previous estimate of the number of people in the region, and

processing at least some of the sensor data according to the varied detection threshold.

13. The method of claim 1

wherein at least one of the plurality of single occupant characteristics is computed based on processing training data by a computer coupled to gather the training data form one or more of the plurality of devices upon satisfaction of a training criterion, and

wherein the training criterion comprises a condition for the satisfaction of the training criterion.

14. The method of claim 2 wherein at least one single occupant characteristic in the single occupancy criterion is a rule indicating one or more combinations of sensor data that cannot be accounted for by a single occupant and thereby defining one or more combinations of sensor data that cannot satisfy the single occupancy criterion.

15. The method of claim 2 wherein the test is passed by the candidate subset of the set of sensor data when each of the plurality of single occupant characteristics are satisfied by the candidate subset of the set of sensor data.

16. The method of claim 2 further comprising:

estimating the number of people in the region according the number of subsets of devices in the set of subsets of devices and a previous estimate of the number of people in the region.

17. A non-transitory computer readable medium containing instructions stored therein for causing a computer processor to perform a method for estimating the number of people in a region of an enclosed space, comprising:

instructions for obtaining sensor data from a plurality of devices within a network of devices, the sensor data being indicative of occupancy within at least a portion of the region,

instructions for obtaining a single occupancy criterion comprising a plurality of single occupant characteristics each being a rule which when satisfied by a candidate subset of the sensor data corresponding to a candidate set of devices from the plurality of devices indicates that the candidate subset of the sensor data can be accounted for by a single occupant,

wherein the single occupancy criterion is satisfied by the candidate subset of the sensor data when a combination of the plurality of single occupant characteristics is satisfied by the candidate subset of the sensor data, wherein the combination is defined in the single occupancy criterion,

instructions for processing one or more subsets of the sensor data according to the single occupancy criterion, to identify a set of subsets of devices such that for each subset of devices in the set of subsets of devices the corresponding subset of the sensor data satisfies the single occupancy criterion, and such that each device in the plurality of devices is included in at least one subset of devices in the set of subsets of devices, and

instructions for estimating the number of people in the region of the enclosed space based on the number of subsets of devices in the set of subsets of devices.

18. The non-transitory computer readable medium of claim 17 further comprising:

instructions for obtaining occupancy data indicative of occupancy in the region,

instructions for evaluating a training criterion on the occupancy data to generate a result indicating the number of people in the region, and based at least in part on the result, obtaining training data from one or more of the plurality of devices,

instructions for processing the training data to generate a single occupant characteristic, and

instructions for adding the single occupant characteristic to the plurality of single occupant characteristics in the single occupancy criterion.

19. The non-transitory computer readable medium of claim 17 further comprising:

instructions for generating the at least one of the plurality of single occupant characteristics by processing training data from the plurality of devices.

20. The non-transitory computer readable medium of claim 17 further comprising:

instructions for estimating the number of people of the region in the enclosed space based on the number of subsets of devices in the set of subsets of devices and based on a previous estimate of the number of people in the region of the enclosed space.

21. A central controller comprising:

at least one processor; and

memory storing computer-readable instructions that, when executed by the at least one processor, cause the central controller to;

obtain sensor data from a plurality of devices, wherein the sensor data from each device of the plurality of devices is indicative of a measure of occupancy within at least a portion of a region in an enclosed space,

access a single occupancy criterion, comprising a test based on a plurality of single occupant characteristics, and which when satisfied by a candidate subset of the sensor data corresponding to a candidate subset of devices from the plurality of devices indicates that the candidate subset of the sensor data can be accounted for by a single occupant, wherein the single occupancy criterion is satisfied for the candidate subset of the sensor data when the test is passed by the candidate subset of the sensor data,

process the single occupancy criterion using one or more subsets of devices from the plurality of devices, to generate a set of subsets of devices such that for each subset of devices in the set of subsets of devices the corresponding subset of the sensor data from all devices in the each subset of devices satisfies the single occupancy criterion, and

estimate the number of people in the region in the enclosed space based on the number of subsets of devices in the set of subsets of devices.

22. The central controller of claim 21 wherein the single occupancy criterion comprises at least one single occupant characteristic that is generated at least in part by processing training data form at least one of the plurality of devices at the central controller.

23. The central controller of claim 21 wherein the computer-readable instructions, when executed by the at least one processor, further cause the central controller to compute the set of subsets of devices by:

iteratively performing, until all devices in the plurality of devices are included in a subset in the set of subsets of devices, the steps of;

a. generating a largest possible subset of devices comprising devices from the plurality of devices that are not already included in a subset of devices in the set of subset of devices and such that the single occupancy criterion is satisfied by the combined sensor data for all devices in the largest possible subset,

b. adding the largest possible subset to the set of subsets of devices.

24. The central controller of claim 21 wherein at least one of the single occupant characteristics is based processing training data by a computer operable coupled to receive the training data from one or more of the plurality of devices.

25. The central controller of claim 21 wherein at least one of the single occupant characteristics is based on processing training data by the central controller.

26. The central controller of claim 21 wherein the central controller is located in a local area network controller and wherein the sensor data for at least some of the plurality of devices is gathered from data packets transmitted by the at least some of the plurality of devices through the local area network controller.

27. A computer implemented method for estimating the number of people in a region of an enclosed space comprising:

obtaining a set of sensor data, indicative of occupancy within at least some of the region, and comprising for each of a plurality of devices within a network of devices corresponding sensor data,

obtaining a single occupancy criterion comprising a plurality of single occupant characteristics,

wherein upon processing at least some of the set of sensor data according to the single occupancy criterion, the single occupancy criterion functions to identify whether the at least some of the set of sensor data can be accounted for by a single occupant,

wherein the single occupancy criterion further comprises a set of rules for the satisfaction of the single occupancy criterion, the set of rules being based at least in part on at least some of the plurality of single occupancy characteristics

processing one or more subsets of the set of sensor data according to the single occupancy criterion, to compute a set of subsets of devices such that for each subset of devices in the set of subsets of devices the corresponding subset of the set of sensor data satisfies the single occupancy criterion, and such that each device in the plurality of devices is included in at least one subset of devices in the set of subsets of devices, and

estimating the number of people in the region of the enclosed space based on the number of subsets of devices in the set of subsets of devices.

28. The method of claim 27 wherein one or more of the set of rules are generated at least in part by processing training data by a computer operably coupled to receive the training data from the network of devices.

29. The method of claim 27 wherein at least one of the set of rules is based on processing training data from a second network of devices located in a second enclosed space.

30. The method of claim 27 further comprising the steps of, prior to computing the set of subsets of devices, modifying one or more of the set of rules based on a previous estimate of the number of people in the region of the enclosed space.