STABILITY: Statistical Verification, Explanation Based Learning, and Interaction Testing for Complex Systems

Complex systems are inherently difficult to analyze because they include multiple sub-systems that interact in stochastic and context-dependent ways. A particular challenge in designing such systems is to foresee the possible emergent interactions between the sub-systems. System repurposing makes the identification of interactions even more difficult, since the interactions vary along with the operational context. SIFT proposes to develop the STABILITY (Statistical Verification, Explanation Based Learning, and Interaction Testing for Complex Systems) system, a complex system design and analysis tool. STABILITY will support model based complex system development. STABILITY will combine human-guided workflow design, automated workflow analysis with statistical verification, succinct workflow analysis summarization with explanation based learning, and systematic workflow suite design with combinatorial interaction testing. STABILITY will provide a simple and effective framework in which to seek out, test, and understand interactions within complex systems.