The Internet of Things (IoT) is rapidly increasing its diffusion, posing great challenges to the research community. IoT systems are composed by smart objects (Things) that are interconnected in order to provide new products and services. The interaction of heterogeneous and distributed smart things guided by software with the physical world brings new sources of safety issues. To this reason, providing valuable and effective solutions to support the verification and validation of such systems is needed. In this paper we introduce a model-driven Thing-In-the-Loop verification and validation approach that transfers the best practices adopted in different embedded system domains towards the IoT world. Starting from models and scenarios representing the structure and behaviors of the IoT system as well as models of its context our approach generates appropriate test cases that are executed in accordance with Model-in-the-Loop, Software-in-the-Loop and Hardware-in-the-Loop techniques. We preliminarily evaluated the feasibility of our approach by applying it in the context of a Smart Mobility case study.
