High-tech system manufacturers are increasingly interested in methodologies to support the servicing of their systems. The growing complexity of these systems, coupled with the criticality of their performance, necessitates advanced tools and techniques for troubleshooting failures and minimizing unexpected downtimes. Additionally, it is challenging to scale service organizations to meet the demands of a rapidly expanding installed base. To address these needs, we propose a model-driven methodology for diagnostics, in which, via a digital diagnostic reasoning system, we leverage design information to assist the service engineer in the troubleshooting process.

A diagnostic model is at the core of the proposed methodology. This model mainly contains information on the functional dependencies between the different subsystems and modules in the system. The structure of the model is based on the system’s functional breakdown, functional exchanges and the deployment on hardware, as they are modelled in Capella. Information about system observables and tests is added to relate field observations to the appropriate system functions. Upon usage, the observations from a troubled system in the field are provided to such a diagnostic model, which then uses the functional dependencies to reason over the possible root-causes.

In this talk, we will show how we use Capella as the main source of information for our diagnostic models and illustrate the use of such a model through a demo. We will also discuss some of the challenges we are still facing and possible next steps.