Advanced identification and classification of sensors and other points in a building automation system

• US 10,001,765 B2
• Filed: 07/01/2016
• Issued: 06/19/2018
• Est. Priority Date: 07/02/2015
• Status: Active Grant

• Current Assignee: CopperTree Analytics Inc.
• Original Assignee: BuildPulse, Inc. (CopperTree Analytics Inc.)
• Inventors: Burt, Jason, Kosnik, Brice
• Primary Examiner(s): Yanchus, III, Paul
• Application Number:
  US15/201,222
• Publication Number:
  US 20170003668A1
• Time in Patent Office: 718 Days
• Field of Search: None
• CPC Class Codes: G05B 15/02 electric
• CPC Class Codes: G05B 19/042 using digital processors G05B19/05 takes precedence
• CPC Class Codes: G05B 2219/25011 Domotique, I-O bus, home automation, building automation
• CPC Class Codes: G05B 2219/2642 Domotique, domestic, home control, automation, smart house

Abstract

A building automation space facility or building automation facility comprising systems and methods for classifying and standardizing sensors and sensor data using time-series data and/or meta data, such as data and meta information available in a control network or hardware is disclosed. The disclosed system and methods provide techniques to rapidly identify and classify sensors using analysis of time-series data (e.g., data collected from a sensor such as by polling the sensor every second, minute, hour, day, month, and so on) and/or meta data (e.g., label names, group names, associated equipment, and so on) from these sensors and associated control systems. In this manner, the disclosed techniques greatly improve the extent to which a building automation facility or system monitors, manages, and reports on various elements within a building or group of buildings.

Claims

1. A method, performed by a building automation space facility of a heating, ventilation, and air-conditioning controller having a processor, for classifying sensors in one or more buildings, the method comprising:

for each of a first plurality of non-classified sensors of the one or more buildings,

retrieving time-series data for the non-classified sensor of the first plurality of non-classified sensors of the one or more buildings,

for each of a plurality of classified sensors, retrieving time-series data for the classified sensor, and

determining a distance between the classified sensor and the non-classified sensor of the one or more buildings,

determining whether a smallest distance determined for the non-classified sensor of the one or more buildings is less than a first predetermined threshold, and

in response to determining that the smallest distance determined for the non-classified sensor of the one or more buildings is less than the first predetermined threshold,

identifying the classified sensor with the smallest determined distance from the non-classified sensor,

retrieving a name for the determined classified sensor with the smallest determined distance,

attributing, to the non-classified sensor of the one or more buildings, the retrieved name for the determined classified sensor with the smallest determined distance from the non-classified sensor, and

controlling the non-classified sensor of the one or more buildings to write the retrieved name for the determined classified sensor to internal storage of the non-classified sensor of the one or more buildings; and

for each of a second plurality of non-classified sensors of the one or more buildings,

retrieving a name for the non-classified sensor of the second plurality of non-classified sensors,

for each of a plurality of names of a classifier name library comprising a plurality of names,

determining a match percentage for the non-classified sensor of the second plurality of non-classified sensors at least in part by comparing the name of the classifier name library to the name retrieved for the non-classified sensor of the second plurality of non-classified sensors,

determining whether the largest match percentage determined for the non-classified sensor of the second plurality of non-classified sensors is greater than a second predetermined threshold, and

in response to determining that the largest match percentage determined for the non-classified sensor of the second plurality of non-classified sensors is greater than the second predetermined threshold,

identifying the name of the classifier name library with the largest determined match percentage for the name retrieved for the non-classified sensor of the second plurality of non-classified sensors,

attributing, to the non-classified sensor of the second plurality of non-classified sensors, the identified name with the largest determined match percentage for the name retrieved for the non-classified sensor, and

controlling the non-classified sensor of the second plurality of non-classified sensors to write the identified name of the classifier name library to internal storage of the non-classified sensor of the second plurality of non-classified sensors.

2. The method of claim 1, wherein the sensors include temperature sensors, Carbon Dioxide sensors, and energy meters.

3. The method of claim 1, wherein the sensors include water flow sensors, equipment status sensors, humidity sensors, and noise sensors.

4. The method of claim 1, wherein each of the first plurality of non-classified sensors is in one room of a first building and wherein each of the plurality of classified sensors is selected from a group of sensors in a different room of the first building.

5. The method of claim 1, wherein determining a match percentage for a first non-classified sensor of the second plurality of non-classified sensors and a first name of the classifier name library comprises:

separating the retrieved name for the first non-classified sensor of the second plurality of non-classified sensors according to a delimiter rule;

separating the first name of the classifier name library into constituent parts according to a delimiter rule; and

comparing the constituent parts of the separated name for the non-classified sensor of the second plurality of non-classified sensors to constituent parts of the first name of the classifier name library.

6. The method of claim 5, wherein comparing the constituent parts of the separated name for the first non-classified sensor of the second plurality of non-classified sensors to constituent parts of the separated first name of the classifier name library comprises:

for each constituent part of the retrieved name for the first non-classified sensor of the second plurality of non-classified sensors,

comparing the constituent part to the constituent parts of the first name of the classifier name library to determine whether the constituent part of the retrieved name for the first non-classified sensor of the second plurality of non-classified sensors matches at least one of the constituent parts of the separated first name of the classifier name library; and

wherein determining a match percentage for the first non-classified sensor comprises;

determining a number of constituent parts of the retrieved name for the first non-classified sensor of the second plurality of non-classified sensors that match at least one of the constituent parts of the separated first name of the classifier name library, and

a total number of constituent parts of the retrieved name for the first non-classified sensor of the second plurality of non-classified sensors.

7. The method of claim 6, wherein each constituent part is a letter.

8. The method of claim 6, wherein at least one of the constituent parts is a word.

9. The method of claim 1, wherein determining the distance between a first classified sensor and a first non-classified sensor comprises:

for each of a plurality of statistical measures,

determining a value of the statistical measure for the first classified sensor,

determining a value of the statistical measure for the first non-classified sensor, and

determining a difference between the determined values of the statistical measure.

10. The method of claim 9, wherein determining the distance between the first classified sensor and the first non-classified sensor further comprises:

for each of the determined differences,

determining the square of the determined difference;

determining the sum of the determined squares; and

determining the square root of the determined sum.

11. The method of claim 1, wherein the second plurality of non-classified sensors is a subset of the first plurality of non-classified sensors.

12. The method of claim 11, further comprising:

for each of a third plurality of non-classified sensors, wherein the third plurality of non-classified sensors is a subset of the second plurality of non-classified sensors,

flagging the non-classified sensor of the third plurality of non-classified sensors for no match.

13. The method of claim 12, wherein each of the plurality of classified sensors is selected from a group of classified local sensors, the method further comprising:

for each of the flagged non-classified sensors of the third plurality of non-classified sensors,

attempting to classify the non-classified sensor based on classified sensors selected from a group of classified account sensors.

14. A building automation space system for classifying points in one or more buildings, the system comprising:

one or more processors;

one or more memories; and

a component configured to, for each of a first plurality of non-classified points of the one or more buildings,

retrieve time-series data for the non-classified point of the first plurality of non-classified points of the one or more buildings,

for each of a plurality of classified points,

retrieve time-series data for the classified point, and

determine a distance between the classified point and the non-classified point,

determine whether a smallest distance determined for the non-classified point of the one or more buildings is less than a first predetermined threshold, and

in response to determining that the smallest distance determined for the non-classified point of the one or more buildings is less than the first predetermined threshold,

identify the classified point with the smallest determined distance from the non-classified point,

retrieve a label for the determined classified point with the smallest determined distance, and

attribute to the non-classified point the retrieved label for the determined classified point with the smallest determined distance from the non-classified point,

wherein the component comprises computer executable instructions stored in the one or more memories for execution by the one or more processors.

15. The building automation space system of claim 14, further comprising:

means for performing a hierarchical classification of one or more of the first plurality of non-classified points; and

means for performing a meta data classification of one or more of the first plurality of non-classified points.

16. The building automation space system of claim 14, wherein at least one of the points is a sensor and wherein at least one of the points is not a sensor.

17. A computer-readable storage medium storing instructions that, if executed by a building automation space system having a processor, cause the building automation space system to perform a method for classifying points in one or more buildings, the method comprising:

for each of a first plurality of non-classified points of the one or more buildings,

retrieving a point name for the non-classified point of the first plurality of non-classified points,

for each of a plurality of classifier names,

determining a match percentage for the non-classified point of the first plurality of non-classified points at least in part by comparing the classifier name to the point name retrieved for the non-classified point of the first plurality of non-classified points,

determining whether the largest match percentage determined for the non-classified point of the first plurality of non-classified points is greater than a first predetermined threshold, and

in response to determining that the largest match percentage determined for the non-classified point of the first plurality of non-classified points is greater than the first predetermined threshold,

identifying a classifier name with the largest determined match percentage for the point name retrieved for the non-classified point, and

attributing, to the non-classified point of the first plurality of non-classified points, the identified classifier name with the largest determined match percentage for the point name retrieved for the non-classified point, and

for each of a second plurality of non-classified points of the one or more buildings,

retrieving time-series data for the non-classified point of the second plurality of non-classified points of the one or more buildings,

for each of a plurality of classified points,

retrieving time-series data for the classified point, and

determining a distance between the classified point and the non-classified point of the one or more buildings,

determining whether a smallest distance determined for the non-classified point of the one or more buildings is less than a second predetermined threshold, and

in response to determining that the smallest distance determined for the non-classified point of the second plurality of non-classified points is less than the second predetermined threshold,

identifying the classified point with the smallest determined distance from the non-classified point of the second plurality of non-classified points,

retrieving a classification for the determined classified point with the smallest determined distance, and

attributing, to the non-classified point of the second plurality of non-classified points, the retrieved label for the determined classified point with the smallest determined distance from the non-classified point of the second plurality of non-classified points.

18. The computer-readable storage medium of claim 17, wherein at least one of the points is a sensor and wherein at least one of the points is not a sensor.

19. A method, performed by a building automation system having a processor, for classifying points in one or more buildings, the method comprising:

for each of a first plurality of non-classified points of the one or more buildings,

retrieving time-series data for the non-classified point of the first plurality of non-classified points of the one or more buildings,

for each of a plurality of classified points,

retrieving time-series data for the classified point, and

determining a distance between the classified point and the non-classified point,

determining whether a smallest distance determined for the non-classified point of the one or more buildings is less than a first predetermined threshold, and

in response to determining that the smallest distance determined for the non-classified point of the one or more buildings is less than the first predetermined threshold,

identifying the classified point with the smallest determined distance from the non-classified point,

retrieving a label for the determined classified point with the smallest determined distance, and

attributing to the non-classified point the retrieved label for the determined classified point with the smallest determined distance from the non-classified point.

20. The method of claim 19, further comprising:

for each of a second plurality of non-classified points of the one or more buildings,

retrieving time-series data for the non-classified point of the second plurality of non-classified points of the one or more buildings,

for each of a plurality of classified points,

retrieving time-series data for the classified point, and

determining a distance between the classified point and the non-classified point of the second plurality of non-classified points of the one or more buildings,

determining whether a smallest distance determined for the non-classified point of the second plurality of non-classified points of the one or more buildings is less than a second predetermined threshold, and

in response to determining that the smallest distance determined for the non-classified point of the second plurality of non-classified points of the one or more buildings is less than the second predetermined threshold,

identifying the classified point with the smallest determined distance from the non-classified point of the second plurality of non-classified points of the one or more buildings,

retrieving a classification for the determined classified point with the smallest determined distance, and

attributing, to the non-classified point of the second plurality of non-classified points of the one or more buildings, the retrieved label for the determined classified point with the smallest determined distance from the non-classified point of the second plurality of non-classified points of the one or more buildings.