AI/ML-powered predictive maintenance for tens of thousands of EVs

AgileEngine creates custom AI and data solutions for an AI startup launched in collaboration with a top-5 luxury car brand featured on the Fortune Global 500 list. The main focus of this collaboration is an ML-as-a-service platform streamlining electric vehicles manufacturing and maintenance. Our team helped ensure that the core AI/ML systems underlying the platform can reliably predict vehicle failures with unparalleled accuracy.

Industries

Automotive, Manufacturing, Big data, Artificial intelligence, Machine learning, Internet of things

Services

Full-stack development, Data engineering, AI engineering, QA and software testing, UX/UI design, UI development

Solutions

Predictive maintenance, AI/ML library, Machine Learning models, Data pipeline, Data visualization

Technologies

Python, AWS, AWS Sagemaker, AWS Glue, MWAA, Redshift, PostgreSQL, Node.js, GraphQL, Serverless Framework, Prisma ORM, DBT, Terraform, TypeScript, React.js, Vite

Outcomes
and highlights

AgileEngine has outperformed the client’s previous remote development vendor, demonstrating higher levels of AI/ML, engineering, and industry expertise
The AI solution we delivered predicts battery failures by analyzing vehicle telemetry, ensuring timely repairs and saving the client up to 95% on component guarantees

Solutions overview

Project overview

Bringing AI to EV manufacturing

The key solution delivered by our AI and data teams is a machine learning library that powers analytics and automation workflows for tens of thousands of electric cars. Based on a vast and textured dataset covering millions of vehicle scenarios, this solution enables our client to improve quality control and optimize repairs.

Key deliverables

Machine learning library capturing and analyzing vehicle data
Pipelines for the extraction, loading, and transformation of millions of vehicle telemetry records
Information management and visualization application
Design and deployment of the infrastructure required by mMachine lLearning models
Data visualization enabling the deep analysis of individual vehicles and automotive fleets
Alert system enabling the prediction of vehicle failures
Built-in analysis tool reducing the manufacturers’ QA time by up to 90%
End-to-end software testing
UX/UI design improvements

Technologies