A qualitative modelling toolbox for control architectures

The authors describe a qualitative modeling toolbox for designing and developing application-independent architectures. The domain chosen to demonstrate the versatility of the work is an advanced robotics workcell that is part of a knowledge-based automation laboratory. Developed out of a remote monitoring application, the toolbox uses object-oriented software, and encapsulates knowledge-based information from a human expert to construct the model. Since it is object-oriented, there is full inheritance between individual objects. The qualitative model can be produced in the form of an interactive user interface that can be used in fault monitoring and diagnosis and report faults. It is noted that using this modular evolvable approach to systems monitoring has meant that one is not simply constrained to diagnosing faults at an architecture's networking level, the workcell level.<
>