METHOD AND APPARATUS FOR AUTOMATION OF A PROGRAMMABLE DEVICE

• US 20110118857A1
• Filed: 12/02/2009
• Published: 05/19/2011
• Est. Priority Date: 12/02/2009
• Status: Active Grant

• Current Assignee: Velvetwire LLC
• Original Assignee: Velvetwire LLC
• Inventors: Bodnar, Eric
• Granted Patent: US 8,849,462 B2
• Time in Patent Office: Days
• US Class Current: 700/47
• CPC Class Codes: G05B 13/026 using a predictor
• CPC Class Codes: G05B 13/0265 the criterion being a learning criterion
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 2219/25377 New sequence as function of deviation from predicted result, state
• CPC Class Codes: G05B 2219/2614 HVAC, heating, ventillation, climate control
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house
• CPC Class Codes: G05D 23/19 characterised by the use of electric means G05D23/1393 takes precedence
• CPC Class Codes: G05D 23/1904 variable in time
• CPC Class Codes: G06N 20/00 Machine learning

Abstract

A method and apparatus for a computer-implemented adaptive automation module comprising an event recorder to store one or more events for a predetermined period, and a timeline pattern generator logic to create a timeline for the predetermined period. The module further comprising marker creator logic to generate a marker to abstract the timeline data from the event data for controlling a device.

Claims

1. A method for automation of a controller comprising:

recording a timeline pattern based on a record of events;

establishing a marker on the timeline pattern based on the record of events;

adjusting the marker to approximate location consistency of the events in the timeline;

computing a weight for the marker that reflects time consistency of the events; and

triggering output based on the marker.

2. The method of claim 1, wherein the creating the timeline pattern further comprises:

archiving events into an event histogram; and

collapsing the event histogram to create the timeline pattern.

3. The method of claim 1, further comprising:

passing the timeline pattern through a low-pass filter, the low-pass filter removing oscillations from the timeline pattern.

4. The method of claim 1, further comprising:

qualifying the marker using external data.

5. The method of claim 4, wherein the external data comprises at least one of:

time of day information, ambient light information, weather information, pricing information, and demand response data.

6. The method of claim 1, wherein the weight for the marker is computed based on a time range of the marker and a number of events within the time range of the marker.

7. The method of claim 1, wherein generating a marker comprises, for each slot in a timeline pattern:

determining whether an activity value for a current slot in the timeline pattern matches an activity value for an subsequent slot in the timeline pattern; and

adding a marker to a list of markers if the activity value for the current slot and the activity value for the next slot do not match.

8. The method of claim 1, wherein adjusting the marker further comprises:

scanning the record of events for at least one event occurring after the marker and prior to a next marker;

recording a time of the at least one event;

computing an average time for the marker based on the time of the at least one event; and

repositioning the marker based on the average time.

9. The method of claim 1, wherein computing the weight for the marker comprises:

scanning the timeline pattern for one or more events occurring within a time range of the marker; and

assigning the weight to the marker based on the proximity of the one or more events to the marker.

10. A computer-implemented adaptive automation module comprising:

an event recorder to store one or more events for a predetermined period;

a timeline pattern generator logic to create a timeline for the predetermined period;

a marker creator logic to generate a marker to abstract the timeline data from the event data for controlling a device.

11. The system of claim 10, wherein the marker creator logic comprises:

a marker generator to generate the marker based on the timeline;

a marker adjuster to adjust the marker based on the one or more events; and

a marker weight assignor to assign a weight to the marker.

12. The system of claim 11, wherein the weight for the marker is computed based on a time range of the marker and a number of events within the time range of the marker.

13. The method of claim 11, wherein the marker generator is to evaluate each slot in the timeline, determine whether the activity value matches an activity value for an subsequent slot in the timeline pattern, and add a marker to a list of markers if the activity value for the current slot and the activity value for the next slot do not match.

14. The method of claim 11, wherein the marker adjustor is further to scan the one or more events for an outlier event occurring near the marker, and computing an adjusted time for the marker based on a time of the outlier event.

15. The method of claim 11, wherein the marker weight assignor is further to identify in-range events occurring within a time range of a marker, and assigning the weight to the marker based on a proximity of the in-range events to the marker.

16. The system of claim 10, wherein the creating the timeline pattern generator logic is further to generate an event histogram based on the one or more events, and to collapse the event histogram to create the timeline

17. The system of claim 10, further comprising a low pass filter to remove oscillations from the timeline.

18. The system of claim 10, wherein the market creator further is to qualify the marker using external data.

19. The system of claim 14, wherein the external data comprises at least one of:

time of day information, ambient light information, weather information, pricing information, and demand response data.

20. A adaptive automation module comprising:

a storage medium;

a processor associated with the storage medium and configured to;

create a timeline pattern based on a record of events;

abstract the timeline data from the record of events by generating an intermediate layer of markers that are adjusted for consistency, thereby reducing effects of variations.