As high-tech systems become more and more capital intensive, there is a strong drive to have no unscheduled down time. On the road towards this ambitious goal lie several milestones, one of which is to keep the amount of down time of such systems below a threshold.

A very important element of this milestone is the amount of time needed to diagnose the root cause of failure. Due to the increasing complexity of high-tech systems, finding the root cause of a failure becomes harder and harder, especially for those failures that have many effects (“collateral damage”) in the system. Such failures produce an avalanche of undesired effects, easily confusing service engineers in finding the actual root cause.

The current way of working of service engineers is mostly trial-and-error and experience-based. A main reason for this is the lack of available knowledge at service side regarding the design of the system. This knowledge traditionally resides in the design and engineering departments. The current trend of digitalization and adoption of model-based system engineering, exposes this knowledge and provides the opportunity of model-based diagnostics. Such a method, however, needs to be scalable and accurate, where accuracy is defined as indicating the correct action plan for an equipment failure.

This project builds upon a series of projects on model-based diagnostics that ESI has done with their industry and academic partners in the past years. The project addresses diagnostics of performance degradation as an extension to the methodologies developed so far, which primarily address diagnostics of system failures. In the project, also extensions to the system failure diagnostics methodology will be developed.