Systems and methods for identifying a vehicle platform using machine learning on vehicle bus data

Patent metadata

Field	Value
Publication number	US-11727271-B2
Application number	US-202016803785-A
Country	US
Kind code	B2
Filing date	Feb 27, 2020
Priority date	Feb 27, 2020
Publication date	Aug 15, 2023
Grant date	Aug 15, 2023

Abstract

Embodiments of the invention include a vehicle telematics system that obtains vehicle bus data for a time period, determines identification information regarding a vehicle platform using a machine learning process on the vehicle bus data, and obtains a set of communication data for communicating with at least one vehicle module on the vehicle bus based on the identified vehicle platform.

First claim

What is claimed is:

1. A method of identifying a vehicle platform using vehicle bus data, comprising:
obtaining vehicle bus data for a time period from a vehicle, wherein the vehicle bus data comprises data communicated via a vehicle bus of the vehicle during the time period;
identifying a vehicle platform using a machine learning process based on the vehicle bus data, wherein the vehicle platform specifies a set of configuration settings for a set of vehicle modules for the vehicle, wherein the set of configuration settings comprises On-Board Diagnostic Parameter IDs (OBD-II PIDs); and
obtaining a set of communication data for communicating with at least one vehicle module on the vehicle bus based on the identified vehicle platform;
wherein identifying the vehicle platform using the machine learning process comprises:
extracting data from a data-field from the vehicle bus data, wherein the data-field comprises vehicle module ID;
determining frequency information for each vehicle module ID, wherein the frequency information comprises a number of occurrences for each vehicle module ID in the vehicle bus data during the time period; and
providing the extracted data and the frequency information to a trained machine learning model that classifies the extracted data and the frequency information to a label indicative of the identified vehicle platform.

2. The method of claim 1 , wherein the vehicle bus is a Controller Area Network (CAN) vehicle bus and the communication data is a set of On-board Diagnostic Parameter IDs (OBD-II PIDs).

3. The method of claim 2 , further comprising:
obtaining information regarding a year, make, and model (YMM) of the identified vehicle platform; and
using the YMM information to obtain a set of OBD-II PIDs for the vehicle.

4. The method of claim 1 , wherein the machine learning process is a supervised neural network model that has been trained on a set of vehicle bus data obtained from a plurality of different vehicles with different YMMs.

5. The method off claim 1 , wherein the machine learning process is a unsupervised machine learning process that performs cluster analysis on vehicle bus data obtained from a plurality of vehicles to group the vehicle bus data.

6. A vehicle telematics device, comprising:
a processor and a memory storing a vehicle telematics application; and
a communication interface for communicating with a remote server system and a plurality of vehicle modules on a vehicle bus of the vehicle;
wherein the processor of the telematics device, on reading the vehicle telematics application, is directed to:
obtain vehicle bus data for a time period, wherein the vehicle bus data comprises data communicated via the vehicle bus during the time period;
identify a vehicle platform using a machine learning process on the vehicle bus data, wherein the vehicle platform specifies a set of configuration settings for a set of vehicle modules for the vehicle, wherein the set of configuration settings comprises On-Board Diagnostic Parameter IDs (OBD-II PIDs); and
obtain a set of communication data for communicating with at least one vehicle module on the vehicle bus based on the identified vehicle platform;

wherein to identify the vehicle platform using the machine learning process comprises to:
extract data from a data-field from the vehicle bus data, wherein the data-field comprises vehicle module ID;
determine frequency information for each vehicle module ID, wherein the frequency information comprises a number of occurrences for each vehicle module ID in the vehicle bus data during the time period; and
provide the extracted data and the frequency information to a trained machine learning model that classifies the extracted data and the frequency information to a label indicative of the identified vehicle platform.

7. The vehicle telematics device of claim 6 , wherein the extracted data and the frequency information are provided to a remote server system that performs a machine learning model on the extracted data and the frequency information.

8. The vehicle telematics device of claim 6 , wherein the vehicle bus is a Controller Area Network (CAN) vehicle bus and the communication data is a set of On-board Diagnostic Parameter IDs (OBD-II PIDs).

9. The vehicle telematics device of claim 8 , wherein the processor of the telematics device, on reading the vehicle telematics application, is further directed to:
obtaining information regarding a year, make, and model (YMM) of the identified vehicle platform; and
using the YMM information to obtain a set of OBD-II PIDs for the vehicle.

10. The vehicle telematics device of claim 6 , wherein the machine learning process is a supervised neural network model that has been trained on a set of vehicle bus data obtained from a plurality of different vehicles with different YMMs.

11. The vehicle telematics device of claim 6 , wherein the machine learning process is a unsupervised machine learning process that performs cluster analysis on vehicle bus data obtained from a plurality of vehicles to group the vehicle bus data.

12. The vehicle telematics device of claim 6 , wherein the time period is dynamically adjusted and determined based on an accuracy of the machine learning process on a set of collected bus data.

Assignees

• Calamp Corp

Inventors

• Abhishek
• Fu Brian
• Rau Amrit Krishna

Classifications

• G06N3/09

  Supervised learning · CPC title

• G06N3/0499

  Feedforward networks · CPC title

• G06N3/08 Primary

  Learning methods · CPC title

• G06F17/18

  for evaluating statistical data {, e.g. average values, frequency distributions, probability functions, regression analysis (forecasting specially adapted for a specific administrative, business or logistic context G06Q10/04)} · CPC title

• G07C5/008

  communicating information to a remotely located station (transmission systems for measured values G08C) · CPC title