MOLTKE : an integrated workbench for fault diagnosis in engineering systems

Abstract

We present the MOLTKE workbench for technical diagnosis, a fully implemented knowledge acquisition environment which has been tested on several real world applications (e.g. fault diagnosis of CNC machining centers). MOLTKE combines different techniques of second generation expert systems (qualitative, case-based, inductive reasoning). Thus, it is able, to process knowledge from the three main sources of technical diagnosis tasks: knowledge about the function and behavior of the machine (technical model), abstract diagnostic knowledge from the expert (mental model), as well as episodic knowledge of service technicians (cases).