On Early Statistical Requirements Validation of Cyber-Physical Space Systems

Abstract

Cyber-physical space systems are becoming increasingly important. Such systems have to satisfy requirements that are heavily affected by the physical space they operate in and by the active entities inhabiting the space, whose dynamic behaviors generate continuous topological changes. Reasoning about requirements in the early design phases is extremely challenging. High-level design can be facilitated by systematic application of separation of concerns throughout modeling, analysis, and early requirements validation. We outline an approach that identifies key recurrent concerns arising in cyber-physical space systems, supports systematic and semi-automatic modeling of separate concerns, and a formally defined composition of the separately developed models. Early requirements validation is then supported by leveraging statistical model checking techniques. We illustrate our approach through an example disaster scenario in a smart city.