Analysis of Truck Compressor Failures Based on Logged Vehicle Data

摘要: Vehicle uptime is getting increasingly important as the transport solutions become more complex and industry seeks new ways of being competitive. Traditional Fleet Management Systems are gradually extended with features to improve reliability, such better maintenance planning. Typical diagnostic predictive methods require extensive experimentation modelling during development. This unfeasible if complete vehicle addressed it would too much engineering resources.This thesis investigates unsupervised supervised for predicting maintenance. The data driven use amounts data, either streamed, on-board or historic aggregated from off-board databases. rely on a telematics gateway that enables vehicles communicate back-office system. Data representations, aggregations models, sent wirelessly an system which analyses deviations. These later associated repair history form knowledge base can be used predict upcoming failures other show same deviations.The further different doing representations deviation detection. first one presented, COSMO, self-organised approach demonstrated fleet city buses. It automatically comes up most interesting uses consensus based isolate deviating vehicle. second outlined super-vised classification earlier collected statistics in label usage statistics. A classifier trained learn patterns precede specific repairs thus method air compressor AB Volvo’s database