PREMA-HEAPS – Predictive Maintenance and Fault Detection for Efficient Heat Pumps

The main goals are to implement machine learning algorithms to detect sub-optimal operation, predict failures, and perform automated root cause analysis. The project will also investigate measurement uncertainty in heat pump data and its impact on fault detection algorithms. The methodology involves conducting state-of-the-art analysis, developing AI/ML algorithms, validating them through real-world data, quantifying measurement uncertainty, and providing guidelines to minimize uncertainty. Expected outcomes include innovative predictive maintenance methods, fault detection algorithms, measurement uncertainty, and industry adoption facilitated through knowledge dissemination activities.