If a fully loaded truck breaks down on the road, for example due to injector damage, the component failure must be repaired, which incurs costs that usually have to be paid by the manufacturer. Late delivery results in convention penalties and a drop in quality ranking, which is bad for follow-up orders.
Based on the telematics data, fault memory entries and repair information, a data set is built to predict failures. The developed algorithm identifies patterns in the ECU data that can be used to distinguish healthy from failed vehicles. With the pattern learned and validated, predictions can be made for all vehicles in the future as to the likelihood of injector failure.
With the current status, 92% of injector failures can be correctly predicted.
This leads to lower warranty costs in the long term, delay penalties are prevented and follow-up orders can be secured.
For this predictive maintenance project, Alexander Thamm GmbH received the Best of Consulting Award 2016 from Wirtschaftswoche.
Are you interested in your own use cases?
Data Science Use Cases
Data & AI Knowledge
Creating joint value from Data & AI