Model checking of security properties: A case study on Human-Robot Interaction processes

Abstract

In Model-Driven Security (MDS) approaches, model checking is a natural way of assuring security properties. It offers a powerful complement to more traditional approaches like testing, and can reduce the overall development costs. In fact, it forces the developers to tackle security issues from the very beginning of their projects, in contrast to common practice where security is often considered just as an afterthought, and indeed quite an expensive one, in many projects and applications. This is especially the case when interesting, potentially disruptive, realms are investigated. Human-Robot Interaction (HRI) and co-working is a hot topic today, with robots becoming every day more pervasive in our lives. Nevertheless, many developers and researchers do not pay enough attention on the incredible consequences that an attack on or a failure of a robot may pose for the human companion or co-worker. In this paper, we take a closer look at security properties for a typical HRI process which has been modeled in DIME, and show how they can be verified with the GEAR model checker.