Software engineering challenges for investigating cyber-physical incidents

Cyber-Physical Systems (CPS) are characterized by the interplay between digital and physical spaces. This characteristic has extended the attack surface that could be exploited by an offender to cause harm. An increasing number of cyber-physical incidents may occur depending on the configuration of the physical and digital spaces and their interplay. Traditional investigation processes are not adequate to investigate these incidents, as they may overlook the extended attack surface resulting from such interplay, leading to relevant evidence being missed and testing flawed hypotheses explaining the incidents. The software engineering research community can contribute to addressing this problem, by deploying existing formalisms to model digital and physical spaces, and using analysis techniques to reason about their interplay and evolution. In this paper, supported by a motivating example, we describe some emerging software engineering challenges to support investigations of cyber-physical incidents. We review and critique existing research proposed to address these challenges, and sketch an initial solution based on a meta-model to represent cyber-physical incidents and a representation of the topology of digital and physical spaces that supports reasoning about their interplay.