To fulfil market demand, the complexity of high-tech industrial systems is increasing every year. As a consequence, diagnosing failures leads to long down times, often because the issue at hand has to be escalated all the way to the development department to access the required knowledge. In this paper we propose a model based diagnostic methodology to support the field service engineer in finding the root cause of an unscheduled downtime in a timely way by bringing actionable design knowledge at the service engineer’s fingertips. We start by modeling the knowledge on the machine hardware decomposition, deployment and functional behavior by using a domain specific language. The language allows for an object oriented description of the system, with the functional behavior captured as probabilistic relationships between hardware components. This system description is then automatically transformed into a Bayesian belief network (BN): the diagnostic model. When a diagnosis is required, we infer the health state of the system’s components by instantiating the BN with discretized sensor and control data collected prior to the occurrence of a downtime. Finally, we compute the root cause hypotheses as the minimum sets of faulty components that are consistent with the evidence. We explain in a visual way the essential part of the diagnostic reasoning by pruning the causal graph underlying the BN, augmented with the data items supporting the hypothesis. We illustrate the methodology on a thermal conditioning subsystem. Our approach is domain independent and applicable to a wider range of cyber physical systems (CPS).
model based diagnostics, bayesian networks, probabilistic reasoning, industrial application
The Prognostic and Health Management Society advocates open-access to scientific data and uses a Creative Commons license for publishing and distributing any papers. A Creative Commons license does not relinquish the author’s copyright; rather it allows them to share some of their rights with any member of the public under certain conditions whilst enjoying full legal protection. By submitting an article to the International Conference of the Prognostics and Health Management Society, the authors agree to be bound by the associated terms and conditions including the following:
As the author, you retain the copyright to your Work. By submitting your Work, you are granting anybody the right to copy, distribute and transmit your Work and to adapt your Work with proper attribution under the terms of the Creative Commons Attribution 3.0 United States license. You assign rights to the Prognostics and Health Management Society to publish and disseminate your Work through electronic and print media if it is accepted for publication. A license note citing the Creative Commons Attribution 3.0 United States License as shown below needs to be placed in the footnote on the first page of the article.
First Author et al. This is an open-access article distributed under the terms of the Creative Commons Attribution 3.0 United States License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.