Method and apparatus for evaluating risk based on sensor monitoring
First Claim
Patent Images
1. A computer program product tangibly stored on a non-transient computer readable hardware storage device, the computer program product for forming predictions of something going wrong with a physical object at a physical premises, the computer program product comprising instructions to cause a processor to:
- collect sensor information from plural groups of sensors deployed in a specific premises;
continually analyze the collected sensor information by executing one or more unsupervised learning models;
produce operational states of sensor information;
produce sequences of state transitions from the operational states of the sensor information;
determine from the sequences of state transitions, normal sensor states and drift sensor states, with respect to the physical object being monitored by the plural groups of sensors, wherein the drift sensor states are indicative of something going wrong with the physical object;
determine a prediction of something going wrong with the physical object based on the drift sensor states;
assign scores to the produced sequences;
determine from the assigned scores a quotation score;
store the sensor information with the determined quotation score in a persistent storage system; and
send the prediction and the quotation score to an external system;
receive a request for a particular line of insurance;
determine from the particular line of insurance, factors and associated sensor data on which to apply the unsupervised learning models; and
generate from results of executing the unsupervised machine learning models the quotation score for a particular line of insurance, where the quotation score is a measure of risk for a particular line of insurance.
7 Assignments
0 Petitions
Accused Products
Abstract
Described are techniques for determining a quotation score that would be applicable across lines of insurance and/or carriers, and which involves the collection in real time of sensor information from plural groups of sensors deployed in a specific premises and associated with intrusion detection, access control, burglar, fire alarm systems and surveillance systems and/or other systems that monitor for physical/chemical/biological conditions. The techniques execute unsupervised learning models to continually analyze the collected sensor information to produce sequences of state transitions that are assign scores and from which a quotation score is produced.
120 Citations
20 Claims
-
1. A computer program product tangibly stored on a non-transient computer readable hardware storage device, the computer program product for forming predictions of something going wrong with a physical object at a physical premises, the computer program product comprising instructions to cause a processor to:
-
collect sensor information from plural groups of sensors deployed in a specific premises; continually analyze the collected sensor information by executing one or more unsupervised learning models; produce operational states of sensor information; produce sequences of state transitions from the operational states of the sensor information; determine from the sequences of state transitions, normal sensor states and drift sensor states, with respect to the physical object being monitored by the plural groups of sensors, wherein the drift sensor states are indicative of something going wrong with the physical object; determine a prediction of something going wrong with the physical object based on the drift sensor states; assign scores to the produced sequences; determine from the assigned scores a quotation score; store the sensor information with the determined quotation score in a persistent storage system; and send the prediction and the quotation score to an external system; receive a request for a particular line of insurance; determine from the particular line of insurance, factors and associated sensor data on which to apply the unsupervised learning models; and generate from results of executing the unsupervised machine learning models the quotation score for a particular line of insurance, where the quotation score is a measure of risk for a particular line of insurance. - View Dependent Claims (2, 3, 4, 5, 6, 7, 8)
-
-
9. The system of 1, wherein the sensors comprise smoke detectors or fire detectors.
-
10. The system of 1, wherein the sensors comprise cameras.
-
11. A system comprises:
-
a server computer comprising processor and memory, the server computer coupled to a network; a storage device storing a computer program product for forming predictions of failure of operation of a physical object at a physical premises, the computer program product comprising instructions to cause the server to; collect sensor information from plural groups of sensors deployed in a specific premises; continually analyze the collected sensor information by executing one or more unsupervised learning models; produce operational states of sensor information; produce sequences of state transitions from the operational states of sensor information; determine from the sequences of state transitions, normal sensor states and drift sensor states, with respect to the physical object being monitored by the plural groups of sensors, the drift sensor states are indicative of something going wrong with the physical object; determine a prediction of something going wrong with the physical object based on the drift sensor states; assign scores to the produced sequences; determine from the assigned scores a quotation score; store the sensor information with the quotation score in a remote persistent storage system; and send the prediction and the quotation score to an external system; receive a request for a particular line of insurance; determine from the particular line of insurance, factors and associated sensor data on which to apply the unsupervised learning models; and generate from results of executing the unsupervised learning models the quotation score for a particular line of insurance, where the quotation score is a measure of risk for a particular line of insurance. - View Dependent Claims (12, 13, 14, 15, 16, 17, 18)
-
-
19. The system of 9, wherein the sensors comprise smoke detectors or fire detectors.
-
20. The system of 9, wherein the sensors comprise cameras.
Specification
-
- Resources
Litigation Campaign Assessment
Thank you for your request. You will receive a custom alert email when the Litigation Campaign Assessment is available.
×
-
Current AssigneeJohnson Controls Tyco IP Holdings LLP (Johnson Controls International plc)
-
Original AssigneeSensormatic Electronics LLC (Johnson Controls International plc)
-
InventorsVavrasek, David
-
Primary Examiner(s)Baird, Edward J
-
Application NumberUS15/147,093Publication NumberTime in Patent Office1,370 DaysField of Search705 4, 705 35- 40US Class CurrentCPC Class CodesG06F 21/577 Assessing vulnerabilities a...G06N 20/00 Machine learningG06Q 10/0635 Risk analysis of enterprise...G06Q 40/08 InsuranceG08B 13/2491 Intrusion detection systems...H04L 63/0428 wherein the data content is...H04L 63/1425 Traffic logging, e.g. anoma...H04W 4/70 Services for machine-to-mac...
