Peer-to-peer Collaborative Vehicle Health Management – the Concept and an Initial Study

Abstract

Advanced vehicle diagnostics and prognostics (D&P) technology enhances ownership experience and reduces corporate warranty cost. D&P performance optimization requires significant algorithm tuning and a large amount of test data collection, which is resource consuming.

In this paper, we propose a novel D&P framework called Collaborative Vehicle Health Management (CVHM) to automatically optimize the D&P algorithms on a host vehicle, using the field data collected from peer vehicles encountered on the road. The carefully designed system architecture and learning algorithms enhance D&P performance without costly human intervention. The experimental results on battery remaining useful life prediction show the effectiveness of the proposed framework. This proposed framework has been implemented in a small test fleet as a proof-of-concept prototype.