Vehicle Testing

AI supports vehicle test planning, test generation, and validation by using historical and specification data to reduce manual engineering work.

Business impact

• validated result accuracy — Nissan reported an initial bolt-tightening study with a validated result of at least 96% true.
• implementation bugs identified — Volvo’s real-world test run surfaced 22 legitimate API implementation bugs.

Data requirements

• historical test data (tabular) — Used by Nissan’s machine-learning platform to support test planning and validation.
• specifications (text) — Volvo’s system interprets specifications to generate test cases.
• vehicle signals (sensor) — Volvo’s workflow matches API properties to vehicle signals.
• real truck API data (text) — The test system operates on real truck APIs and their implementation details.

AI methods and techniques

• machine learning — Used on historical test data to support planning and validation of physical vehicle tests.
• LLM-based test automation — Used to interpret specifications, match API properties to vehicle signals, generate test cases, and run tests.
• automated test generation — Used to turn specifications and signals into executable test cases.

Industries

Real-world evidence

2 documented case studies on record.

Companies using this: Nissan, Volvo.