AnnMarie Ericsson, TP:
Utilizing Existing Analysis Tools in Development of Predictable sets of
This thesis proposal addresses support for utilizing a
heterogeneous set of analysis tools when predictable rule-based
systems are developed. One of the main problems of the rule-based
paradigm is that it is hard to analyze the behavior of rule sets,
which is conflicting with the high predictability requirements
typically associated with real-time systems. Further, analysis
tools developed for rule-based systems typically address a
specific platform or a specific part of the development chain.
In our approach, rules are initially specified in a high-level
language. % managing cooperation between platforms and tools.
We enable powerful analysis tools, not necessary designed for rule
based development, to be utilized for analyzing the rule base.
This is done by transforming the set of rules, with maintained
semantics, to a representation suitable for the target analysis
tool. Our approach supports developers to take advantage of a
variety of analysis tools when rule-based systems are developed.
Initial research has focused on utilizing the model checker in
Uppaal for formal verification of the rule set.