Predictive Maintenance in Automotive

Assignment: Deployment of predictive maintenance using artificial intelligence and machine learning algorithms for one of the customer’s production lines (paint shop).

Input data

As the input data for the project served both historically measured data and expert knowledge (service protocols).

Our solution

Anomaly Detector (location, time, type)

• identification of unusual changes and unknown behavioral patterns, based on automatic earning from typical behavior
• methods used: positional and correlation characteristics

The result

Prediction of a timeframe for the next breakdown and time of appropriate service intervention.