Predictive maintenance of vehicle components

Patent metadata

Field	Value
Publication number	US-12093901-B2
Application number	US-202016951706-A
Country	US
Kind code	B2
Filing date	Nov 18, 2020
Priority date	Aug 25, 2020
Publication date	Sep 17, 2024
Grant date	Sep 17, 2024

Abstract

A method for predictive maintenance of a component of a vehicle includes collection of first operational data, from a telematics device of the vehicle, for a set of parameters associated with the component. From the first operation data, a first dataset is selected based on an engine status of an engine of the vehicle. The first dataset is processed to obtain a plurality of feature values and the plurality of feature values are segregated into a plurality of clusters. A classifier is trained based on the plurality of clusters to determine the health status of the component. Real-time or near real-time operational data for the set of parameters is received, from the telematics device of the vehicle. The received operational data is used as an input to the trained classifier and the health status of the component is determined based on an output of the trained classifier.

First claim

What is claimed is:

1. A method for predictive maintenance of a component of a vehicle, the method comprising:
collecting, by a server, from a telematics device of the vehicle, first operational data for a set of parameters associated with the component of the vehicle over a first time-interval, and wherein the component of the vehicle is one of a battery or an alternator;
selecting, by the server, from the first operational data, a first dataset based on an engine status of an engine of the vehicle, wherein
the first dataset is collected by the server at a time the engine has an OFF status,
the selected first dataset is associated with the battery of the vehicle, and
the selected first dataset includes a discharge voltage of the battery during the OFF status of the engine of the vehicle;

processing, by the server, the first dataset to obtain a plurality of feature values;
segregating, by the server, the plurality of feature values into a plurality of clusters, wherein
the segregating is based on clustering algorithms using machine learning models to label the plurality of clusters into one of a first cluster, a second cluster, and a third cluster, wherein each of the first cluster, the second cluster, and the third cluster of the plurality of clusters corresponds to a health status of the component of the vehicle, and

the health status of the component of the vehicle corresponds to a working state and performance of the vehicle;
training, by the server, a classifier of the server based on the plurality of clusters, wherein the training of the classifier comprises determining the health status of the component;
outputting, by the trained classifier, the health status of the component based on historical maintenance data of the component of the vehicle;
validating, by the server, an accuracy level of the classifier based on the health status output by the trained classifier and the historical maintenance data, wherein the accuracy level of the classifier is validated to generate a validation output;
re-training, by the server, the classifier based on the validation output;
receiving, by the server, from the telematics device of the vehicle, second operational data of the component of the vehicle, wherein
the second operational data corresponds to real-time or near real-time operational data for the set of parameters associated with the component, and
the real-time or near real-time operational data is input to the re-trained classifier;

determining, by the server, the health status of the component of the vehicle based on an output of the re-trained classifier for the real-time or near real-time operational data; and
determining, by the server, a remaining useful life of the component of the vehicle for a second time-interval, wherein the remaining useful life of the component is determined based on the health status determined for the real-time or near real-time operational data.

2. The method of claim 1 , wherein the set of parameters includes the engine status of the engine, voltage data associated with the component, one or more diagnostic trouble codes associated with the component, and a past count of each of the one or more diagnostic trouble codes.

3. The method of claim 1 , wherein the set of parameters includes one of an engine ON/OFF count of the engine, a maximum and minimum voltage associated with the component, an average voltage associated with the component, a deviation of minimum voltage associated with the component, an average minimum voltage associated with the component, an average maximum voltage associated with the component, and an average engine ON time of the engine.

4. The method of claim 1 , further comprising training, by the server, the classifier based on vehicle data of the vehicle, wherein the vehicle data includes one of a vehicle model, a make of the vehicle, a vehicle manufacturing year, and a city of operation of the vehicle.

5. The method of claim 1 , wherein the validation output is used as a feedback to improve the accuracy level of the trained classifier.

6. The method of claim 1 , wherein the plurality of feature values includes one of a plurality of moving average values, a plurality of moving standard deviation values, a plurality of cumulative sum values, a plurality of rolling mean values, or a plurality of rolling standard deviation values obtained by processing the first dataset.

Assignees

• Ani Tech Private Ltd

Inventors

• Mahipal Punit
• Hinduja Hitesh
• Voruganti Vamsee Krishna
• Agarwal Gaurav

Classifications

• G07C5/0808

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

• G05B23/0283 Primary

  Predictive maintenance, e.g. involving the monitoring of a system and, based on the monitoring results, taking decisions on the maintenance schedule of the monitored system; Estimating remaining useful life [RUL] (preventive maintenance, i.e. planning maintenance according to the available resources without monitoring the system G06Q10/06) · CPC title

• G06N20/00

  Machine learning · CPC title

• B60R16/0234

  related to maintenance or repairing of vehicles · CPC title

• G07C5/006

  Indicating maintenance · CPC title