Prediction of false alarms in sensor-based security systems

• US 10,359,771 B2
• Filed: 06/08/2017
• Issued: 07/23/2019
• Est. Priority Date: 06/08/2017
• Status: Active Grant

• Current Assignee: Tyco Fire & Security GmbH (Johnson Controls International plc)
• Original Assignee: Tyco Fire & Security GmbH (Johnson Controls International plc)
• Inventors: Trainor, Christine, Subramanian, Gopi, Vavrasek, David
• Primary Examiner(s): Bousono, Orlando
• Application Number:
  US15/617,259
• Publication Number:
  US 20180356809A1
• Time in Patent Office: 775 Days
• CPC Class Codes: G05B 23/0229 knowledge based, e.g. expert systems; genetic algorithms
• CPC Class Codes: G05B 23/0283 Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06
• CPC Class Codes: G06F 16/00 Information retrieval; Database structures therefor; File system structures therefor
• CPC Class Codes: G08B 29/04 Monitoring of the detection circuits
• CPC Class Codes: G08B 29/14 checking the detection circuits
• CPC Class Codes: G08B 29/185 Signal analysis techniques for reducing or preventing false alarms or for enhancing the reliability of the system
• CPC Class Codes: H04W 84/18 Self-organising networks, e.g. ad-hoc networks or sensor networks

Abstract

Techniques for detecting changes in operational characteristics of a group of sensor devices are described. The techniques collect sensor information from plural sensor devices deployed in a system, with the collected sensor information including sensor data and sensor device metadata. The techniques continually analyze the collected sensor information to detect changes in the operational characteristics of a sensor device in the group of sensor devices. Upon detection of changes in the operational characteristics of the sensor, a database that stores maintenance organization contact information is accessed and a request is generated for maintenance on the sensor device, which request is sent to the maintenance organization.

Claims

1. A non-transitory computer program product tangibly stored on a non-transitory computer readable hardware storage device, the computer program product for detecting changes in operational characteristics of a group of sensor devices deployed in a system, the computer program product comprising instructions to cause a processor to:

collect sensor information from the group of sensor devices with the collected sensor information including sensor operational data and sensor device metadata;

execute one or more unsupervised learning models to continually analyze the collected sensor information to detect changes in operational characteristics of a sensor device in the group of sensor devices, with the changes correlated to an imminent potential failure of the sensor device;

produce sequences of sensor state transitions;

detect during the continual analysis of the sensor data that the detected changes in the operational characteristics correspond to one or more of the sequences of sensor state transitions being a drift sequence, by correlating the detected drift sequence to a stored determined condition to determine whether a potential false alarm could occur by the sensor device being in the drift sequence;

upon detection of a change in the operational characteristics of the sensor device that would indicate the potential failure or occurrence of a potential false alarm, access a database that stores maintenance organization contact information;

generate based on the detected changes and the access to the database a request for maintenance on the sensor device; and

send the request to the maintenance organization contact.

2. The non-transitory computer program product of claim 1 wherein the sensor data is collected continuously.

3. The non-transitory computer program product of claim 1 wherein for the produced sequences of sensor state transitions the instructions determine from state transition metrics that are stored in a state transition matrix, occurrence of a drift condition for the sensor device in the drift condition;

determine from the sensor device metadata, age of the sensor device in the drift condition; and

access a database by sensor device type to retrieve metadata on sensor devices of a type of the one or more sensor devices in the drift condition, and having an age approximate to the determined age of the sensor device.

4. The non-transitory computer program product of claim 1 wherein for the produced sequences of sensor state transitions the instructions generate a second request for maintenance based on the determined drift condition and the age of the sensor device, with the second request for maintenance being for one or more of the sensor devices in the group in the detected drift condition, and for those sensor devices having the approximate age.

5. A system comprising:

a gateway to couple sensor information from plural sensor devices to a network;

a server computer comprising processor and memory, the server computer coupled to the network;

a storage device storing a computer program product for detecting conditions at a premises, the computer program product comprising instructions to cause the server computer to;

collect the sensor information from the plural sensor devices, with the collected sensor information including sensor data and sensor device metadata;

execute one or more unsupervised learning models to continually analyze the collected sensor information to produce operational state of sensor information;

produce sequences of sensor state transitions;

detect during the continual analysis of sensor data that one or more of the sequences of sensor state transitions corresponds to one sensor device being in a drift sequence, by correlating the drift sequence to a stored determined condition to determine whether a potential false alarm could occur by the sensor device being in the drift sequence;

access a database that stores maintenance organization contact information; and

generate an alert based on the drift sequence and the access to the database with the alert being a request for maintenance on the one sensor devices.

6. The system of claim 5 wherein the computer program product, further comprises instructions to:

convert states corresponding to sensor device values into a semantic representation of a state; and

assign a label to the semantic representation of the state.

7. The system of claim 5 wherein the system monitors a premises and is installed at the premises and further comprises:

plural sensor devices installed at the premises.

8. The system of claim 5 wherein the system monitors a premises and is remote from the premises.

9. The system of claim 5 wherein the sensor data is collected continuously.

10. The system of claim 5 wherein for the produced sequences of sensor state transitions the instructions:

determine from state transition metrics that are stored in a state transition matrix, occurrence of a drift condition for one or more of the sensor devices;

determine from the sensor device metadata, age of the one or more sensor devices in the drift condition; and

access a database by sensor device type to retrieve metadata on sensor devices of a type of the one or more sensor devices in the drift condition, and having an age approximate to the determined age of the sensor device.

11. The system of claim 10 wherein for the produced sequences of sensor state transitions the instructions generate a second set of alerts based on the determined drift condition and the age of the sensor device, a request for maintenance on the one or more of the sensor devices in the determined drift condition, and for those sensor devices having the approximate age.

12. A computer implemented method comprises:

collecting sensor information from plural sensor devices deployed in a system, with the collected sensor information including sensor data and sensor device metadata;

executing one or more unsupervised learning model to continually analyze the collected sensor information to produce operational states of sensor information;

producing sequences of sensor state transitions;

detecting during the continual analysis of sensor data that one or more of the sequences of sensor state transitions corresponds to one sensor device being in a drift sequence, by correlating the detected drift sequence to a stored determined drift sequence condition to determine whether a potential false alarm could occur by the sensor device being in the detected drift sequence;

determining from the sensor device metadata, age of the one sensor device in the detected drift sequence;

accessing a database by sensor device type to retrieve metadata on sensor devices of a type of the one sensor device in the detected drift sequence, and having an age approximate to the determined age of the sensor device;

accessing a database that stores maintenance organization contact information; and

generating an alert based on the detected drift sequence and the access to the database that stores the maintenance organization contact information, with the alert being a request for maintenance on the one sensor device.

13. The method of claim 12, further comprising:

converting states corresponding to sensor device values into a semantic representation of a state; and

assigning a label to the semantic representation of the state.

14. The method of claim 12, further comprising providing the system at a premises.

15. The method of claim 14, wherein the plural sensor devices are installed at the premises.

16. The method of claim 12, wherein the system monitors a premises and is remote from the premises.

17. The method of claim 12, wherein the sensor data is collected continuously.

18. The method of claim 12, wherein for the sequences of sensor state transitions, the method further comprises:

determining from state transition metrics that are stored in a state transition matrix, occurrence of a drift condition for one or more of the sensor devices.

19. The method of claim 12, wherein for the produced sequences of sensor state transitions, the method further comprises:

providing a second set of alerts based on the determined drift condition and the age of the sensor device, a request for maintenance on the one or more of the sensor devices in the determined drift condition, and for those sensor devices having the approximate age.

20. The method of claim 12, wherein the sensor devices are cameras.