ANALYSIS OF VEHICLE DATA TO PREDICT COMPONENT FAILURE

• US 20160035150A1
• Filed: 07/30/2014
• Published: 02/04/2016
• Est. Priority Date: 07/30/2014
• Status: Active Grant

• Current Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Original Assignee: Verizon Patent and Licensing Incorporated (Verizon Communications Inc.)
• Inventors: Barfield, James Ronald Jr., Welch, Stephen Christopher, Barbier, Marc Edmond
• Granted Patent: US 9,881,428 B2
• Time in Patent Office: Days
• US Class Current: 1/1
• CPC Class Codes: G05B 23/0254 based on a quantitative model, e.g. mathematical relationships between inputs and outputs; functions: observer, Kalman filter, residual calculation, Neural Networks
• CPC Class Codes: G07C 5/008 communicating information to a remotely located station transmission systems for measured values G08C
• CPC Class Codes: G07C 5/0841 Registering performance data recording measured values G01D; information storage G11B

Abstract

Vehicle data may be analyzed to predict potential component failures, diagnostic trouble codes (DTCs), or other mechanical failures relating to the vehicle. In one implementation the vehicle data may be received from a number of vehicles, the vehicle data including DTCs generated by on-board diagnostic (OBD) systems of the vehicles. The vehicle data may be evaluated using a predictive model to output predictions of DTCs that are likely to occur for a particular vehicle.

Claims

1. A method comprising:

receiving, by one or more computing devices, vehicle data from a plurality of vehicles, the vehicle data including diagnostic trouble codes (DTCs) generated by on-board diagnostic (OBD) systems of the plurality of vehicles;

evaluating, by the one or more computing devices, the vehicle data according to a predictive model, to output predictions of DTCs that are likely to occur for a particular one of the plurality of vehicles;

generating, by the one or more computing devices and based on the output predictions of the DTCs, a report describing potential maintenance issues for the particular one of the plurality of vehicles; and

transmitting, by the one or more computing devices, the report to an entity associated with the particular one of the plurality of vehicles.

2. The method of claim 1, wherein the vehicle data additionally includes:

sensor data generated by telematics devices associated with the plurality of vehicles;

information describing vehicle specific information of the plurality of vehicles;

or

information describing repairs performed on the plurality of vehicles.

3. The method of claim 2, wherein the vehicle specific information includes one or more of:

a vehicle type of the plurality of vehicles;

engine types associated with the plurality of vehicles;

or

recall information associated with vehicle types of the plurality of vehicles.

4. The method of claim 1, the method further comprising:

generating the predictive model, based on the vehicle data, using supervised machine learning techniques or unsupervised machine learning techniques.

5. The method of claim 4, further comprising:

generating the predictive model using the supervised machine learning techniques when a number of available examples of the vehicle data for vehicles that have experienced the predicted DTCs is above a threshold; and

generating the predictive model using the unsupervised machine learning techniques when the number of available examples of the vehicle data for the vehicles that have experienced the predicted DTCs is below the threshold.

6. The method of claim 1, further comprising:

clustering the vehicle data, to obtain driver archetypes relating to driving patterns of drivers of the plurality of the vehicles;

classifying the drivers of the plurality of vehicles into the obtained driving archetypes; and

obtaining the output predictions of the DTCs based on the classification of the drivers into the obtained driving archetypes.

7. The method of claim 6, wherein the predictions of the DTCs include an estimate of the probability of DTC occurring within a specific time period.

8. The method of claim 1, further comprising:

determining, based on the output predictions from the predictive model, one or more components, associated with the particular vehicle, that are likely to fail.

9. The method of claim 8, further comprising:

determining predictions relating to when the one or more components are likely to fail.

10. A computing system comprising processing circuitry to:

receive vehicle data from a plurality of vehicles, the vehicle data including diagnostic trouble codes (DTCs) generated by on-board diagnostic (OBD) systems of the plurality of vehicles;

evaluate the vehicle data according to a predictive model to output predictions of DTCs that are likely to occur for a particular one of the plurality of vehicles;

generate, based on the output predictions of the DTCs, a report describing potential maintenance issues for the particular one of the plurality of vehicles; and

transmit the report to an entity associated with the particular one of the plurality of vehicles.

11. The computing system of claim 10, wherein the vehicle data additionally includes:

sensor data generated by telematics devices associated with the plurality of vehicles;

information describing vehicle specific information of the plurality of vehicles;

or

information describing repairs performed on the plurality of vehicles.

12. The computing system of claim 10, wherein the processing circuitry is further to:

generate the predictive model, based on the vehicle data, using supervised machine learning techniques or unsupervised machine learning techniques.

13. The computing system of claim 12, wherein the processing circuitry is further to:

generate the predictive model using the supervised machine learning techniques when a number of available examples of the vehicle data for vehicles that have experienced the predicted DTCs is above a threshold; and

generate the predictive model using the unsupervised machine learning techniques when the number of available examples of vehicle data for the vehicles that have experienced the predicted DTCs is below the threshold.

14. The computing system of claim 10, wherein the processing circuitry is further to:

cluster the vehicle data, to obtain driver archetypes relating to driving patterns of drivers of the plurality of the vehicles;

classify the drivers of the plurality of vehicles into the obtained driving archetypes; and

obtain the output predictions of the DTCs based on the classification of the drivers into the obtained driving archetypes.

15. The computing system of claim 10, wherein the processing circuitry is further to:

determine, based on the output predictions from the predictive model, one or more components, associated with the particular vehicle, that are likely to fail.

16. A computing system comprising processing circuitry to:

receive vehicle data relating to a vehicle, the vehicle data being received from a telematics device associated with the vehicle and the vehicle data including;

diagnostic trouble codes (DTCs) generated by an on-board diagnostic (OBD) system, and

data from sensors associated with the vehicle;

input the vehicle data to a predictive model trained to output maintenance issues that are likely to occur for the vehicle, the predictive model operating to classify a vehicle use pattern associated with the vehicle into one of a plurality of patterns and to generate the output maintenance issues based on the vehicle use pattern; and

transmit an indication of the output of the predictive model, to an entity associated with the vehicle, when the output of the predictive model indicates maintenance issues are likely to occur for the vehicle.

17. The computing system of claim 16, wherein the data from the sensors includes information relating to acceleration of the vehicle, speed of the vehicle, motor revolutions per unit time of the vehicle, engine load of the vehicle, intake air temperature associated with the vehicle.

18. The computing system of claim 16, wherein the data from the sensors includes processed values that are based on the data from the sensors, the processed values including indications of vehicle hard stops, hard turns, or fast accelerations.

19. The computing system of claim 16, wherein the received vehicle data additionally includes:

information describing a vehicle type;

information describing repairs performed on the plurality of vehicles;

information describing vehicle services of the plurality of vehicles;

or

information describing vehicle maintenance records of the plurality of vehicles.

20. The computing system of claim 16, wherein the predictive model is generated on a per manufacturer and per model basis of the vehicle.

21. The computing system of claim 16, wherein the predictive model is generated using unsupervised machine learning techniques.

22. The computing system of claim 16, wherein the computing system is implemented as part of the telematics device associated with the vehicle or as part of the vehicle, and wherein the predictive model is received from a remote server via a wireless interface.

23. The computing system of claim 16, wherein the maintenance issues, output by the predictive model, includes predicted DTCs, likely vehicle repairs, or vehicle component failures.

24. The computing system of claim 16, wherein the processing circuitry is further to:

input the vehicle data to a second predictive model, trained based on supervised machine learning techniques, the second predictive model to output additional maintenance issues that are likely to occur for the vehicle.