Sensor packs that are configured based on business application

• US 10,425,702 B2
• Filed: 09/30/2015
• Issued: 09/24/2019
• Est. Priority Date: 09/30/2015
• Status: Active Grant

• Current Assignee: Johnson Controls Tyco IP Holdings LLP (Johnson Controls International plc)
• Original Assignee: Sensormatic Electronics LLC (Johnson Controls International plc)
• Inventors: Trainor, Christine, Vavrasek, David
• Primary Examiner(s): Sherwin, Ryan W
• Assistant Examiner(s): Kingston, Shawna M
• Application Number:
  US14/870,493
• Publication Number:
  US 20170094376A1
• Time in Patent Office: 1,455 Days
• Field of Search: None
• CPC Class Codes: G06Q 10/00 Administration; Management
• CPC Class Codes: G06Q 40/08 Insurance
• CPC Class Codes: G08B 31/00 Predictive alarm systems characterised by extrapolation or other computation using updated historic data
• CPC Class Codes: H04Q 2209/823 where the data is sent when the measured values exceed a threshold, e.g. sending an alarm
• CPC Class Codes: H04Q 9/00 Arrangements in telecontrol or telemetry systems for selectively calling a substation from a main station, in which substation desired apparatus is selected for applying a control signal thereto or for obtaining measured values therefrom

Abstract

Techniques for detecting physical conditions at a physical premises from collection of sensor information from plural sensors execute one or more unsupervised learning models to continually analyze the collected sensor information to produce operational states of sensor information, produce sequences of state transitions, detect during the continual analysis of sensor data that one or more of the sequences of state transitions is a drift sequence, correlate determined drift state sequence to a stored determined condition at the premises, and generate an alert based on the determined condition. Various uses are described for these techniques.

Claims

1. A computer program product tangibly stored on a non-transitory computer readable hardware storage device, the computer program product for determining a sensor set for a given type of premises, the computer program product comprising instructions to cause a processor to:

collect sensor information from a plurality of groups of different types of sensors deployed in a plurality of different types of premises, with the sensor information including sensor data, sensor identity information and premises identification information;

convert the sensor information for each sensor into semantic representations of a state of a corresponding premises;

produce operational states from the sensor information for the plurality of different types of premises;

produce sequences of state transitions from the operational states;

execute one or more unsupervised learning models that analyze the semantic representations and the operational states to determine from the sensor information for each of the groups of different types of sensors a prediction of an immediate next operational states;

for a prediction of an unsafe operational state,

determine empirically, characteristics of each of the plurality of different types of premises occurring during a determined unsafe operational state;

produce alerts based on the determined unsafe operational state;

determine from the operational states which of the sensors in the groups of different types of sensors contributed to the alerts for each of the plurality of different types of premises and for a given type of premises selected from the plurality of different types of premises,

determine a set of sensors for use in detecting conditions at other premises of a type corresponding to the given type of premises.

2. The computer program product of claim 1, wherein the computer program product further comprises instructions to cause the processor to:

classify the set of sensors as a sensor pack for the premises type corresponding to the given type of premises.

3. The computer program product of claim 2 wherein a specific configuration of the sensor pack is determined for a business segment for the given type of premises.

4. The computer program product of claim 1 wherein for each of the plurality of different types of premises, the program determines which sensor information is more relevant to detection of specific anomalies than other sensor information according to execution of a machine learning unsupervised algorithms.

5. The computer program product of claim 1 wherein the computer program receives sensor information from the groups of sensors for a specific type of the plural premises and the computer program product further comprises instructions to cause the processor to:

determine which of the groups of sensors are relevant for specific types of equipment; and

determine which of the groups of sensors are relevant for monitoring the plurality of different types of premises.

6. A system comprises:

plural groups of sensors, each group comprised of plural different types of sensors to sense plural different physical chemical and optical properties, with the plural groups of sensors installed at a plurality of different types of premises;

a gateway at each premises to couple each of the plural groups of sensors to a network;

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

a non-transitory storage device storing a computer program product for detecting conditions at each of the plurality of different types of premises, the computer program product comprising instructions to cause the server to;

collect sensor information from the plural groups of sensors, with the collected sensor information including sensor data, sensor identity information and premises identification information;

convert the collected sensor information for each sensor into a semantic representations of a state of a corresponding premises;

produce operational states from the sensor information for the plurality of different types of premises;

produce sequences of state transitions from the operational states;

assign labels to the produced operational states;

execute one or more unsupervised learning models that analyze the semantic representations and the produced operational states to determine for each of the different types of premises from the collected sensor information for each of the plural groups of sensor devices, a prediction of an immediate next operational state;

for a prediction of an unsafe operational state,

determine empirically, characteristics of each premises occurring during a determined unsafe operational state;

produce alerts based on the determined unsafe operational state;

determine from produced operational states which sensors in the plural groups of sensors contributed to the produced alerts for each of the different types of premises; and

for a given type of premises selected from the plurality of different types of premises,

determine a set of sensors for use in detecting conditions at other premise of a type corresponding to the given type of premises.

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

classify the determined set of sensors as a sensor pack for the premises type corresponding to the given type of premises.

8. The system of claim 7 wherein a specific configuration of sensor packs is determined for a business segment for the given type of premises.

9. The system of claim 6 wherein for each business segment the program determines which sensor information is more relevant to detection of specific anomalies than other sensor information according to execution of a machine learning unsupervised algorithms.

10. The system of claim 6 wherein the computer program receives sensor information from a plural groups of sensors for a specific instance of the plurality of different types of premises, and the computer program product further comprises instructions to:

determine which of the plural groups of sensors are relevant for specific types of equipment; and

determine which of the plural groups of sensors are relevant for monitoring the plurality of different types of premises.

11. A computer implemented method comprises:

collecting sensor information from a plural groups of sensors that are comprised of plural different types of sensors, and which groups of sensors are deployed in a plurality of different types of premises, with the collected sensor information including sensor data, sensor identity information and premises identification information;

sending by each of plural gateways, the collected sensor information to one or more server computers that comprise processor devices and memory;

converting the collected sensor information for each sensor into a semantic representations of a state of a corresponding premises;

producing operational states from the sensor information for the plurality of different types of premises;

producing by the one or more server computers sequences of state transitions from the operational states;

assigning labels to the produced operational states;

executing by the one or more server computers, one or more unsupervised learning models that analyze the semantic representations and the produced operational states to determine for the plurality of different types of premises from the collected sensor information for each of the plural groups of sensors, a prediction of an immediate next operational state;

for a prediction of unsafe operational state,

determining empirically, characteristics of each of the plurality of different types of premises occurring during a determined unsafe operational states;

producing, by the one or more server computers, alerts based on the determined unsafe operational states;

determining by the one or more server computers from the produced operational states which of the plural groups of sensors contribute to the produced alerts for the plurality of different types of premises; and

for a given premises type selected from the plurality of different types of premises,

determining by the one or more server computers, a set of sensors for use in detecting conditions at other premises of a type corresponding to the given premises type.

12. The method of claim 11 further comprising:

classifying by the one or more server computers the determined set of sensors as a sensor pack for the premises type corresponding to the given premises type.

13. The method of claim 12 wherein a specific configuration of sensor packs is determined for a business segment for the given premises type.

14. The method of claim 11, further comprising:

determining by the one or more server computers for the plurality of different types of premises, which sensor information is more relevant to detection of specific anomalies than other sensor information according to execution of a machine learning unsupervised algorithms.

15. The method of claim 11 further comprising:

determining by the one or more server computers which of the plural groups of sensors are relevant for specific types of equipment; and

determining by the one or more server computers which of the plural groups of sensors are relevant for monitoring the plurality of different types of premises.