Quick Planner for Verification and Validation of Distributed Hybrid Systems

With all the uncertainties in complex Distribute Hybrid Systems (DHS), one key probelm is the definition of a reasonable set of tests (and tools) that will evaluate both the extrmees and the nominal behaviors associated with real-world situations. This general lack of confidence for AI components has prevented a wider acceptance of KBS use in DHS. The challenge is how to test the logic, rules, algorithms and inference engines in the context of traditional S/W systems. The goal of this SBIR is to address the issues associated with V&V/testing of DHS combining conventional and KBS software. We propose to provide designers, developers and testers with methodologies and tools to support their activities. To accomplish this goal we will: 1) Develop formal methodologies for integrating KBS with conventional software, with specific emphasis on the process of V&V/testing of DHS. 2) Develop a data base of V&V tools that can be commercialized. We will also become uniquely qualified for providing independent V&V/testing to systems incorporating KBS.