Automated Risk Scenario Generation Using System Functional and Structural Knowledge
Simulation may be the most practical way to assess the risk of systems with complex behaviors such as those that include hardware, software and human elements. However, since under normal design conditions human-designed systems generally perform in familiar and expected ways, a typical simulation will frequently lead to known and anticipated results. As such, the simulation program wastes a lot of time on familiar results without generating new knowledge about the system’s vulnerabilities. In order to increase our knowledge of risk, it would be preferable to push the system toward its limits to test the system’s ability to handle more difficult situations. Such an approach can help system designers to better understand risky situations and close the vulnerability gaps in their design. The primary objective of this study is to develop a risk simulation Planner (SimpraPlan) which generates scenarios that can explore the system’s vulnerabilities and offer a superior assessment of the risks involved. The Planner uses high level engineering knowledge (including the functional requirements and physical structure of the system) to generate scenarios that can exploit the system’s vulnerabilities. In this paper, the scenario generation process is explained in detail and scenarios generated by the SimpraPlan are compared with those generated by classical approaches to risk assessment.