Technical University of Munich, Department of Computer Science
MQM Paper: [SachenbacherMalikStruss 98]


Martin Sachenbacher, Andreas Malik, Peter Struss

From Electrics to Emissions: Experiences in Applying Model-based Diagnosis to Real Problems in Real Cars
 

Working Papers of the 9th International Workshop on Principles of Diagnosis (DX-98), Sea Crest Resort, Cape Cod, MA, USA, pp. 246-253, 1998.

Also in: Working Papers of the ECAI-98 Workshop on Model-based Systems and Qualitative Reasoning, Brighton, UK, 1998.

Abstract

This paper reports on experiences and preliminary results gained in applying model-based diagnosis on diesel engine management systems in passenger vehicles. It states some major requirements that have been identified in this application. In particular, it is crucial to reason across different physical domains, to exploit emission-related symptoms like increased carbon emissions in the exhaust gas for identifying failures in upstream subsystems. The main foundation of our approach is to use qualitative deviation models, which serve as a coherent modeling paradigm for the different domains. Second, the models are used in a state-based diagnosis framework, which checks consistency of the model with observed states only, to avoid simulation of the system's behavior over time. We present preliminary results of putting this approach into practice for real problems in demonstrator vehicles. The evaluation provides empirical evidence for the utility of model-based techniques to improve the discriminative power of automated vehicle diagnosis both in on-board and off-board situations. The experiments also raised a number of open issues that are subject to further research.

Download


If you have comments or suggestions, email me at mauss@in.tum.de
created: Jakob Mauss, April 3, 1998, last updated: Jakob Mauss, December 4, 1998