Interpreting presence signals using historical data
A method includes obtaining historical event data for events detected over a past period of time by sensors within a property, receiving a set of current event data for one or more events detected by one or more of the sensors within the property, determining that the set of current event data matches a pattern of events indicated by the historical event data, generating, based on the pattern of events, a confidence score for the set of current event data, wherein the confidence scores reflects a confidence that a person is not within the property, determining that the confidence score satisfies a confidence threshold associated with an action to be performed when a person is not within the property, and triggering execution of the action.
1. A computer-implemented method comprising:
detecting, using first sensor data for a first monitoring system at a first property, one or more events at the first property;
detecting, using the first sensor data for the one or more events, a pattern of recurring events;
determining, using the first sensor data for the pattern of recurring events, that a first confidence score for the pattern of recurring events does not satisfy a confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
detecting, using second sensor data for two or more second monitoring systems each of which is at a different physical location, the one or more events, wherein the second sensor data includes the first sensor data for the first monitoring system at the first property;
detecting, using the second sensor data for the one or more events, the pattern of recurring events;
determining, using the second sensor data for the pattern of recurring events, that a second confidence score for the pattern of recurring events satisfies the confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
storing, in computer memory, data for the pattern of recurring events for use by the first monitoring system to detect the pattern of recurring events using sensor data from one or more sensors for the first property and perform an action.
2. The method of claim 1 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system at a second property and the first monitoring system at the first property, the one or more events.
3. The method of claim 1 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system in a vehicle and the first monitoring system in a building at the first property, the one or more events.
4. The method of claim 1 , comprising:
determining, for the pattern of recurring events, two or more automated actions to perform when the pattern of recurring events is detected;
determining, for each of the two or more automated actions, a corresponding action threshold that indicates, given a likelihood that the pattern of recurring events has been detected, a threshold that the likelihood must satisfy for a system to perform the corresponding automated action; and
storing, in computer memory, data for the two or more automated actions and the respective action thresholds.
5. The method of claim 1 , wherein detecting the one or more events comprises detecting at least one of one or more positive events for an activity at the first property, or one or more negative events for a lack of activity at the first property.
6. The method of claim 1 , wherein detecting the pattern of recurring events comprises:
identifying groups of sensor events using cluster analysis;
identifying unusual sensor events using anomaly detection; and
identifying dependencies using association rule mining.
7. The method of claim 1 , wherein detecting the one or more events comprises detecting the one or more events using sensor data from one or more of a camera, a door sensor, a motion sensor, a phone, a contact sensor, a vehicle sensor, a glass break sensor, an environmental sensor, a temperature sensor, a water sensor, a rain sensor, a wind sensor, a light sensor, a smoke sensor, a carbon monoxide detector, an air quality sensor, a health monitoring sensor, or a radio-frequency identification sensor.
8. A system comprising one or more computers and one or more storage devices on which are stored instructions that are operable, when executed by the one or more computers, to cause the one or more computers to perform operations comprising:
detecting, using first sensor data for a first monitoring system at a first property, one or more events at the first property;
detecting, using the first sensor data for the one or more events, a pattern of recurring events;
determining, using the first sensor data for the pattern of recurring events, that a first confidence score for the pattern of recurring events does not satisfy a confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
detecting, using second sensor data for two or more second monitoring systems each of which is at a different physical location, the one or more events, wherein the second sensor data includes the first sensor data for the first monitoring system at the first property;
detecting, using the second sensor data for the one or more events, the pattern of recurring events;
determining, using the second sensor data for the pattern of recurring events, that a second confidence score for the pattern of recurring events satisfies the confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
storing, in computer memory, data for the pattern of recurring events for use by the first monitoring system to detect the pattern of recurring events using sensor data from one or more sensors for the first property and perform an action.
9. The system of claim 8 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system at a second property and the first monitoring system at the first property, the one or more events.
10. The system of claim 8 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system in a vehicle and the first monitoring system in a building at the first property, the one or more events.
11. The system of claim 8 , the operations comprising:
determining, for the pattern of recurring events, two or more automated actions to perform when the pattern of recurring events is detected;
determining, for each of the two or more automated actions, a corresponding action threshold that indicates, given a likelihood that the pattern of recurring events has been detected, a threshold that the likelihood must satisfy for a system to perform the corresponding automated action; and
storing, in computer memory, data for the two or more automated actions and the respective action thresholds.
12. The system of claim 8 , wherein detecting the one or more events comprises detecting at least one of one or more positive events for an activity at the first property, or one or more negative events for a lack of activity at the first property.
13. The system of claim 8 , wherein detecting the pattern of recurring events comprises:
identifying groups of sensor events using cluster analysis;
identifying unusual sensor events using anomaly detection; and
identifying dependencies using association rule mining.
14. The system of claim 8 , wherein detecting the one or more events comprises detecting the one or more events using sensor data from one or more of a camera, a door sensor, a motion sensor, a phone, a contact sensor, a vehicle sensor, a glass break sensor, an environmental sensor, a temperature sensor, a water sensor, a rain sensor, a wind sensor, a light sensor, a smoke sensor, a carbon monoxide detector, an air quality sensor, a health monitoring sensor, or a radio-frequency identification sensor.
15. A non-transitory computer storage medium encoded with instructions that, when executed by one or more computers, cause the one or more computers to perform operations comprising:
detecting, using first sensor data for a first monitoring system at a first property, one or more events at the first property;
detecting, using the first sensor data for the one or more events, a pattern of recurring events;
determining, using the first sensor data for the pattern of recurring events, that a first confidence score for the pattern of recurring events does not satisfy a confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
detecting, using second sensor data for two or more second monitoring systems each of which is at a different physical location, the one or more events, wherein the second sensor data includes the first sensor data for the first monitoring system at the first property;
detecting, using the second sensor data for the one or more events, the pattern of recurring events;
determining, using the second sensor data for the pattern of recurring events, that a second confidence score for the pattern of recurring events satisfies the confidence threshold that indicates that the pattern of recurring events will likely be observed in the future;
storing, in computer memory, data for the pattern of recurring events for use by the first monitoring system to detect the pattern of recurring events using sensor data from one or more sensors for the first property and perform an action.
16. The computer storage medium of claim 15 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system at a second property and the first monitoring system at the first property, the one or more events.
17. The computer storage medium of claim 15 , wherein detecting, using the second sensor data, the one or more events comprises detecting, using the second sensor data for a second monitoring system in a vehicle and the first monitoring system in a building at the first property, the one or more events.
18. The computer storage medium of claim 15 , the operations comprising:
determining, for the pattern of recurring events, two or more automated actions to perform when the pattern of recurring events is detected;
determining, for each of the two or more automated actions, a corresponding action threshold that indicates, given a likelihood that the pattern of recurring events has been detected, a threshold that the likelihood must satisfy for a system to perform the corresponding automated action; and
storing, in computer memory, data for the two or more automated actions and the respective action thresholds.
19. The computer storage medium of claim 15 , wherein detecting the one or more events comprises detecting at least one of one or more positive events for an activity at the first property, or one or more negative events for a lack of activity at the first property.
20. The computer storage medium of claim 15 , wherein detecting the pattern of recurring events comprises:
identifying groups of sensor events using cluster analysis;
identifying unusual sensor events using anomaly detection; and
identifying dependencies using association rule mining.