Methods and systems for data analytics

• US 9,418,493 B1
• Filed: 04/30/2015
• Issued: 08/16/2016
• Est. Priority Date: 04/30/2015
• Status: Active Grant

• Current Assignee: The Boeing Co.
• Original Assignee: The Boeing Co.
• Inventors: Dong, Jian
• Primary Examiner(s): Butler, Rodney
• Application Number:
  US14/700,333
• Time in Patent Office: 474 Days
• US Class Current: 1/1
• CPC Class Codes: G05B 2219/37534 Frequency analysis
• CPC Class Codes: G05B 23/0221 Preprocessing measurements, e.g. data collection rate adjustment; Standardization of measurements; Time series or signal analysis, e.g. frequency analysis or wavelets; Trustworthiness of measurements; Indexes therefor; Measurements using easily measured parameters to estimate parameters difficult to measure; Virtual sensor creation; De-noising; Sensor fusion; Unconventional preprocessing inherently present in specific fault detection methods like PCA-based methods
• CPC Class Codes: G05B 23/0235 based on a comparison with predetermined threshold or range, e.g. "classical methods", carried out during normal operation; threshold adaptation or choice; when or how to compare with the threshold

Abstract

A method includes receiving a stream of sensor data regarding operating conditions of a vehicle, and processing first and second data sets of the stream of sensor data in time and frequency domains. The method also includes identifying differences within and between processing results from processing the first and second data sets, and determining that at least one difference in the identified differences is greater than or equal to one or more predefined thresholds. Further, the method includes, based on determining the at least one difference in the identified differences being greater than or equal to the one or more predefined thresholds, recording one or more processing results corresponding to the identified at least one difference in a data index map (DIM), and analyzing contents of the DIM to evaluate operating conditions of the vehicle.

Claims

1. A method implemented by a computing device for data stream analytics comprising:

controlling a sensor to acquire sensor data regarding operating conditions of a vehicle;

receiving a stream of the sensor data;

processing, by the computing device, a first data set of the stream of the sensor data in a time domain and in a frequency domain;

processing, by the computing device, a second data set of the stream of the sensor data in the time domain and in the frequency domain, wherein the first data set and the second data set are adjacent data sets of the stream of the sensor data;

identifying, by the computing device, one or more of differences within first processing results from processing the first data set, differences within second processing results from processing the second data set, or differences between the first processing results and the second processing results;

determining, by the computing device, that at least one difference in the identified differences is greater than or equal to one or more predefined thresholds;

based on determining the at least one difference in the identified differences being greater than or equal to the one or more predefined thresholds, recording, by the computing device, one or more processing results corresponding to the identified at least one difference in a data index map (DIM); and

analyzing contents of the DIM to evaluate operating conditions of the vehicle.

2. The method of claim 1, wherein processing the first data set includes calculating a first power spectrum density (PSD) of the first data set, wherein processing the second data set includes calculating a second PSD of the second data set, and wherein determining that at least one difference in the identified differences is greater than or equal to the one or more predefined thresholds includes determining that a difference between the first PSD and the second PSD is greater than or equal to a predefined PSD difference threshold.

3. The method of claim 1, wherein processing the first data set in the time domain includes calculating a first variance between a previous data set of the stream of the sensor data and the first data set, wherein processing the second data set in the time domain includes calculating a second variance between the first data set and the second data set, and wherein determining that at least one difference in the identified differences is greater than or equal to the one or more predefined thresholds includes identifying that a difference between the first variance and the second variance is greater than or equal to a predefined variance difference threshold.

4. The method of claim 1, wherein processing the first data set in the time domain includes identifying first peak data points and first valley data points in the first data set, wherein processing the second data set in the time domain includes identifying second peak data points and second valley data points in the second data set, and wherein determining that at least one difference in the identified differences is greater than or equal to the one or more predefined thresholds includes identifying that a difference between adjacent peak data points in the first peak data points, adjacent valley data points in the first valley data points, adjacent peak data points in the second peak data points, or adjacent valley data points in the second valley data points is greater than or equal to a predefined difference threshold.

5. The method of claim 1, wherein the one or more predefined thresholds are one or more first predefined thresholds, and wherein the method further comprises:

determining, by the computing device, that at least one difference in the identified differences is greater than or equal to one or more second predefined thresholds, wherein the one or more second predefined thresholds indicate a greater identified difference than the one or more first predefined thresholds; and

based on determining the at least one difference in the identified differences being greater than or equal to the one or more second predefined thresholds, generating data for providing an audible and/or visual alert that real-time action relating to operation of the vehicle is required.

6. The method of claim 1, further comprising:

generating data for displaying one or more of the first processing results, the identified differences in the first processing results, the second processing results, the identified differences in the second processing results, or the identified differences between the first processing results and the second processing results.

7. The method of claim 1, wherein the vehicle is an aircraft, and further comprising, based on the evaluated operating conditions of the vehicle, at least one of updating a maintenance schedule of the vehicle, determining operating efficiency of the vehicle, determining a travel route of the vehicle, or investigating an accident or a component failure.

8. The method of claim 1, wherein data points in the stream of the sensor data are associated with timestamps, and further comprising, based on determining the at least one difference in the identified differences being greater than or equal to the one or more predefined thresholds, recording, in the DIM, timestamps associated with data points that correspond to the at least one difference that is greater than or equal to the one or more predetermined thresholds.

9. A system for data stream analytics comprising:

a plurality of memory units including data stream memory (DSM) units configured as a pathway for a data stream to a permanent data storage, input data memory (IDM) units configured as input data buffers, output data memory (ODM) units configured as output data buffers, and a first set of data processing result (DPR) memory units and a second set of DPR memory units, each of the first set of DPR memory units and the second set of DPR memory units configured to store processing results; and

a processor communicatively coupled with the plurality of memory units and configured to;

provide a data stream through the DSM units to the permanent data storage;

copy a first data set of the data stream from the DSM units to the IDM units;

generate a first processing result of the first data set using predefined analytics algorithms;

move the first processing result to the first set of DPR memory units;

copy a second data set of the data stream from the DSM units to the IDM units;

generate a second processing result of the second data set using the predefined analytics algorithms;

move the first processing result in the first set of DPR memory units to the second set of DPR memory units;

move the second processing result to the first set of DPR memory units;

identify that at least one difference within the first processing result, within the second processing result, or between the first processing result and the second processing result is greater than or equal to a predefined threshold; and

record one or more processing results corresponding to the identified at least one difference in a data structure.

10. The system of claim 9, wherein the first data set and the second set are adjacent data sets in the data stream, and wherein each of the first data set and the second data set has the same numbers of data points.

11. The system of claim 9, wherein a total number of the DSM units is greater than or equal a total number of the IDM units, wherein the plurality of memory units and the processor are configured such that a time needed to provide the data stream through the DSM units to the permanent data storage is greater than or equal to an analytics time, and wherein the analytics time is a time duration starting when a particular data set is copied into the IDM units and ending when processing results corresponding to the particular data set are moved out the ODM.

12. The system of claim 9, wherein the processor is configured to copy the first data set from the DSM units to the IDM units in parallel with providing the data stream through the DSM units to the permanent storage, wherein the processor is configured to copy the second data set from the DSM units to the IDM units in parallel with providing the data stream through the DSM units to the permanent storage, and wherein the processor is configured to move the first processing result in the first set of DPR memory units to the second set of DPR memory units in parallel with generating the second processing result.

13. The system of claim 9, wherein the processor is further configured to generate data for displaying one or more of the first processing result, the second processing result, or the identified at least one difference.

14. The system of claim 9, wherein the predefined analytics algorithms include processing in a time domain and processing in a frequency domain.

15. The system of claim 9, wherein the data stream includes sensor data regarding operating conditions of a vehicle, and wherein data points in the data stream are associated with timestamps.

16. A system for data stream analytics comprising:

a plurality of memory units including data stream memory (DSM) units configured as a pathway for a data stream to a permanent data storage, input data memory (IDM) units configured as input data buffers, output data memory (ODM) units configured as output data buffers, and a first set of data processing result (DPR) memory units and a second set of DPR memory units, each of the first set of DPR memory units and the second set of DPR memory units configured to store processing results; and

a processor including at least three processor cores, wherein the processor is communicatively coupled with the memory units and configured to;

provide a data stream through the DSM units to the permanent data storage;

copy a first data set of the data stream and a second data set of the data stream from the DSM units to the IDM units;

generate, using a first processor core, a first processing result of the first data set using predefined analytics algorithms;

generate, using a second processor core, a second processing result of the second data set using the predefined analytics algorithms;

move the first processing result to the first set of DPR memory units;

move the second processing result to the second set of DPR memory units;

identify, using a third processor core, that at least one difference within the first processing result, within the second processing result, or between the first processing result and the second processing result is greater than or equal to a predefined threshold; and

record one or more processing results corresponding to the identified at least one difference in a data structure.

17. The system of claim 16, wherein the first data set and the second set are adjacent data sets in the data stream, and wherein each of the first data set and the second data set has the same numbers of data points.

18. The system of claim 16, wherein a total number of the DSM units is greater than or equal a total number of the IDM units, wherein the plurality of memory units and the processor are configured such that a time needed to provide the data stream through the DSM units to the permanent data storage is greater than or equal to an analytics time, and wherein the analytics time is a time duration starting when a particular data set is copied into the IDM units and ending when processing results corresponding to the particular data set are moved out the ODM.

19. The system of claim 16, wherein the processor is configured to copy the first data set and the second data set from the DSM units to the IDM units in parallel with providing the data stream through the DSM units to the permanent storage.

20. The system of claim 16, wherein the predefined analytics algorithms include processing in a time domain and processing in a frequency domain, wherein the data stream includes sensor data regarding operating conditions of a vehicle, and wherein data points in the data stream are associated with timestamps.