Vehicle fault prediction

Patent metadata

Field	Value
Publication number	US-12417658-B2
Application number	US-202218046233-A
Country	US
Kind code	B2
Filing date	Oct 13, 2022
Priority date	Oct 13, 2022
Publication date	Sep 16, 2025
Grant date	Sep 16, 2025

Abstract

A computer includes a processor and a memory, and the memory stores instructions executable by the processor to receive time-series data from sensors of a vehicle, determine an efficiency of the vehicle, and determine a probability of a fault occurring in the vehicle based on the time-series data and on the efficiency. The efficiency is energy consumption by the vehicle per distance traveled by the vehicle.

First claim

What is claimed is:

1. A computer comprising a processor and a memory, the memory storing instructions executable by the processor to:
receive time-series data from sensors of a vehicle;
determine an efficiency of the vehicle, the efficiency being energy consumption by the vehicle per distance traveled by the vehicle;
in response to an indication of a completion of a trip of the vehicle, determine a plurality of probabilities of a plurality of respective faults occurring in the vehicle based on the time-series data and on the efficiency, the memory storing a plurality of thresholds for the respective faults, the thresholds having different values dependent on respective types of the respective faults;
modify operation of a propulsion system of the vehicle based on the probability of the fault; and
output a message in response to at least one of the probabilities exceeding the respective threshold.

2. The computer of claim 1 , wherein the indication is one of turning off the vehicle or shifting the vehicle to park.

3. The computer of claim 1 , wherein the instructions further include instructions to determine a trip-level metric applicable only to the single trip as a whole, and determining the probabilities is based on the trip-level metric.

4. The computer of claim 1 , wherein the instructions further include instructions to determine the probabilities by executing a machine-learning program, and the time-series data and the efficiency are inputs to the machine-learning program.

5. The computer of claim 4 , wherein the machine-learning program includes a feature-extraction program, and the time-series data is an input to the feature-extraction program.

6. The computer of claim 5 , wherein the machine-learning program includes a classification program, and an output of the feature-extraction program is an input to the classification program.

7. The computer of claim 6 , wherein the efficiency is an input directly to the classification program.

8. The computer of claim 6 , wherein the instructions further include instructions to determine a trip-level metric either applicable only to the single trip of the vehicle as a whole or indicating a difference between a state of the vehicle at a beginning of the trip and at the completion of the same trip, and the trip-level metric is an input directly to the classification program.

9. The computer of claim 4 , wherein the machine-learning program is trained on ground-truth faults and training data, the training data including training time-series data and training efficiencies, the training data being paired to the ground-truth faults.

10. The computer of claim 9 , wherein the ground-truth faults are diagnostic trouble codes.

11. The computer of claim 1 , wherein the time-series data includes a temperature of a vehicle component.

12. The computer of claim 1 , wherein the time-series data includes a tire pressure of a tire of the vehicle.

13. The computer of claim 1 , wherein the types of the faults correspond to respective diagnostic trouble codes.

14. The computer of claim 1 , wherein the instructions to output the message include instructions to at least one of output the message to a screen of the vehicle, transmit the message to a mobile device, or transmit the message to a server remote from the vehicle.

15. The computer of claim 1 , wherein the instructions further include instructions to determine a trip-level metric indicating a difference between a state of the vehicle at a beginning of the trip and the completion of the same trip, and determining the probabilities is based on the trip-level metric.

16. A method comprising:
receiving time-series data from sensors of a vehicle;
determining an efficiency of the vehicle, the efficiency being energy consumption by the vehicle per distance traveled by the vehicle;
in response to an indication of a completion of a trip of the vehicle, determining a plurality of probabilities of a plurality of respective faults occurring in the vehicle based on the time-series data and on the efficiency, the memory storing a plurality of thresholds for the respective faults, the thresholds having different values dependent on respective types of the respective faults;
modifying operation of a propulsion system of the vehicle based on the probability of the fault; and
outputting a message in response to at least one of the probabilities exceeding the respective threshold.

Assignees

• Ford Global Tech Llc

Inventors

• Wolf Chris
• Chase Nicholas B
• Degner Michael W

Classifications

• G07C5/0808 Primary

  Diagnosing performance data (testing of vehicles G01M17/00; testing of electrical installation on vehicles G01R31/005) · CPC title

• G07C5/0841

  Registering performance data (recording measured values G01D; information storage G11B) · CPC title

• G01C21/3469

  Fuel consumption; Energy use; Emission aspects · CPC title

• B60W50/0098

  Details of control systems ensuring comfort, safety or stability not otherwise provided for · CPC title

• B60W2050/0031

  Mathematical model of the vehicle · CPC title