Environment for Evolutionary Automotive Diagnosis

The objective of the project is to design a new generation automotive diagnosis infrastructure that will facilitate the diagnosis of modern vehicles by allowing car makers to:

• collect and analyse data from the vehicles on the road, with a distribution of processing between on-board and off-board segments in order to compute failure rates of vehicle components and identify fault patterns for vehicle models,
• enhance the diagnostic procedures along the life of a particular vehicle model or for future models,
• shorten the delay between the occurrence of fault patterns on a vehicle model and the definition of the corrective action to be deployed in the organization for this vehicle model,
• perform remote troubleshooting by making vehicle diagnostic functions available through wireless links and the Internet,
• and eventually detect more quickly and more accurately the root causes of failures, minimise the vehicle down-time and repair cost.

The project is tentatively defining an AMI-C compliant architecture distributed between on-board and off-board segments that shall use the techniques of distributed objects and mobile Java code. The project is also investigating new methods and techniques to enhance the diagnosis procedures from the analysis of collected data.