CAISR+: Predictive maintenance and machine learning

The CAISR+ project is about using recent machine learning methods and working with real industrial data (streaming data, onboard data, or data in databases) related to predictive maintenance. There are currently two main application tracks:

• usage characterization through machine activity recognition (forklift trucks and heavy-duty trucks)
• modelling survivability of vehicle subsystems (turbochargers, coolant pumps, electric drive batteries).

In addition, CAISR+ includes participating in international conferences and organizing thematic workshops to establish CAISR internationally as an actor in this field. CAISR+ also includes advanced courses on the topic directed at professionals.

Ongoing subprojects in CAISR+ are predicting turbocharger failures, modelling coolant pump failures, detecting “no fault found”, transfer learning from high-frequency to low-frequency databases, machine activity recognition from CAN data, survivability for electromobility batteries, modelling bus usage and battery health. The machine learning topics covered here are transfer learning, self-supervised feature learning, domain adaptation, variational autoencoders, meta learning and explainability.

The CAISR+ project has an industrial advisory board to keep the research relevant for the industrial partners and to point at important industrial developments. Our work on machine learning for predictive maintenance has so far resulted in 19 journal papers and 14 conference papers (2020–2022). There have also been an industrial demonstrator and the transfer of code from academia to industry.