Predicting a future state of a built environment

• US 10,572,834 B2
• Filed: 06/06/2015
• Issued: 02/25/2020
• Est. Priority Date: 06/06/2015
• Status: Active Grant

• Current Assignee: Building Robotics, Inc. (Siemens AG)
• Original Assignee: Enlighted, Inc. (Siemens AG)
• Inventors: Mohan, Tanuj, Krishnamurthy, Vijayalakshmi
• Primary Examiner(s): Rivera Vargas, Manuel A
• Assistant Examiner(s): Perez Bermudez, Yaritza H
• Application Number:
  US14/732,704
• Publication Number:
  US 20160356633A1
• Time in Patent Office: 1,725 Days
• CPC Class Codes: G01D 21/02 Measuring two or more variables by means not covered by a single other subclass
• CPC Class Codes: G06Q 10/04 Forecasting or optimisation specially adapted for administrative or management purposes, e.g. linear programming or "cutting stock problem" market predictions or forecasting for commercial activities G06Q30/0202
• CPC Class Codes: G06Q 10/06 Resources, workflows, human or project management; Enterprise or organisation planning; Enterprise or organisation modelling

Abstract

Apparatuses, methods and systems for predicting a future state of a built environment with a sensor network are disclosed. One sensor network includes a plurality of sensors of a built environment and a controller. Each sensor is operative to generate a sense signal. The controller is operative to collect sense values that represent sense signals of the plurality of sensors, wherein the sense values are collected over time, classify each of the collected sense values based on a type of space of the built environment associated with a corresponding one of the plurality of sensors that generated the sense value, based on a timing of the sense value, and based on other available information of the built environment, predict a future state of each of the plurality of sensors based at least in part on the classifications of each of the sense values, and communicate the future state.

Claims

1. A sensor network, comprising:

a plurality of sensors of a built environment, each sensor operative to generate a sense signal;

an HVAC (heating, ventilation and air conditioning) unit of the built environment;

a controller, the controller operative to;

collect sense values that represent sense signals of the plurality of sensors, wherein the sense values are collected over time;

classify each of the collected sense values based on a type of space of the built environment associated with a corresponding one of the plurality of sensors that generated the sense value, based on a timing of the sense value, and based on other available information of the built environment, wherein classifying each of the collected sense values comprises;

weighting the type of space of the built environment associated with a corresponding one of the plurality of sensors that generated the sense value, weighting the timing of the sense value, and weighting each of a plurality of the other available information of the built environment; and

classifying each of the collected sense values based on the weighting of the type of space of the built environment associated with a corresponding one of the plurality of sensors that generated the sense value, the weighting of the timing of the sense value, and the weighting of each of the plurality of the other available information of the built environment;

predict a future state of each of the plurality of sensors based at least in part on the classifications of each of the sense values;

communicate the future state; and

control the HVAC based on the future state.

2. The sensor network of claim 1, wherein the other available information includes information of a space of the built environment that is different than a space of the built environment that is associated with the corresponding one of the plurality of sensors that generated the sense value.

3. The sensor network of claim 2, wherein the other available information includes sensed occupancy of the space that is different that the space of the built environment that is associated with the corresponding one of the plurality of sensors that generated the sense value.

4. The sensor network of claim 1, wherein the weighting is adaptive.

5. The sensor network of claim 4, wherein the adaptive weighting is adaptively updated over time based on observations of an accuracy of past predicted future states.

6. The sensor network of claim 4, wherein the adaptive weighting is updated over time based on:

a determined accuracy of the predicted state based on the type of space of the built environment associated with a corresponding one of the plurality of sensors that generated the sense value;

a determined accuracy of the predicted state based on the timing of the sense value; and

a determined accuracy of the predicted state based a plurality of the other available information of the built environment.

7. The sensor network of claim 1, wherein predicting the future state of each of the plurality of sensors based at least in part on the classifications of each of the sense values comprises an analysis of the type of space associated with the sensor, an identifier of the sensor, a present time, and a predicted time.

8. The sensor network of claim 1, wherein the controller is further operative to select at least a sub-plurality of the plurality of sensors.

9. The sensor network of claim 8, wherein the controller is further operative to classify each of the collected sense values based on the selected at least the sub-plurality of the plurality of sensors.

10. The sensor network of claim 8, wherein the controller is further operative to predict a future state of the sub-plurality of the plurality of sensors.

11. The sensor network of claim 8, wherein the other available information of the built environment includes sensed values of a group of sensors of a neighborhood group, wherein the group of sensors of the neighborhood group are located adjacent to the sub-plurality of sensors, and wherein the group of sensors of the neighboring group have past sensed values that are correlated with past predicted future states of the sub-plurality of sensors by more than a threshold.

12. The sensor network of claim 11, wherein determining that the group of sensors of the neighboring group have past sensed values that are correlated with past predicted future states of the sub-plurality of sensors by more than a threshold comprises identifying that a predicted state based on the sensed values of the neighboring group is observed to be accurate greater than a threshold percent of time.

13. The sensor network of claim 8, wherein the other available information of the built environment includes sensed values of a group of sensors of a influencing group, wherein the group of sensors of the influencing group are not located adjacent to the sub-plurality of sensors, and wherein the group of sensors of the influencing group have past sensed values that are correlated with past predicted future states of the sub-plurality of sensors by more than a threshold.

14. The sensor network of claim 13, wherein determining that the group of sensors of the influencing group have past sensed values that are correlated with past predicted future states of the sub-plurality of sensors by more than a threshold comprises identifying that a predicted state based on the sensed values of the influencing group is observed to be accurate greater than a threshold percent of time.

15. The sensor network of claim 1, wherein the other available information of the built environment includes at least one of badging data, calendar event data, outside environment data, equipment sign-in log data.

16. The sensor network of claim 1, wherein the future state includes a probability of a location of each sensor being occupied or not occupied.

17. The sensor network of claim 16, wherein the future state includes a probability of locations of the sub-plurality of sensor being occupied or not occupied.

18. The sensor network of claim 16, wherein the future state includes an estimated level of occupancy, wherein the level of occupancy is estimated based on correlating predicted states of the sub-plurality of the plurality of sensors with a secondary occupancy level source.

19. The sensor network of claim 1, wherein the controller is further operative to predict energy consumption based on the future state.

20. The sensor network of claim 1, wherein the controller is further operative to predict space utilization based on the future state.

21. The sensor network of claim 1, wherein the controller is further operative to influence future predicted states base upon feedback of a determination of a determined accuracy of the predicted state.

22. The sensor network of claim 1, wherein the controller is further operative to perform training, including tracking a set of known states over time, wherein the controller is further operative to influence future predicted states based the training.

23. The sensor network of claim 1, wherein each sensor includes a sensor controller, and wherein the sensor controller is operative to:

collect the sense values that represent sense signals of the plurality of sensor, wherein the sense values are collected over time;

classify each of the collected sense values based on a type of space of the sensor, and based on a timing of the sense value;

predict a future state of the sensor; and

communicate the future state.