Models@run.time for Creating In-Cloud Dynamic Cyber-Physical Ecosystems

The applications of Cyber-Physical Systems (CPSs) in large scale, mobile and distributed systems, such as transportation and healthcare systems, foster the development of novel cross-CPS applications. Services, in such applications, enable the emergence of multiple end-to-end cyber-physical scenarios, formed dynamically based on their demands, e.g., Disaster recovery systems. This calls for new distributed service composition models that integrate different CPS services from multiple application domains for a given purpose -- CPS ecosystems. The in-Cloud availability of CPSs can significantly improve the development process of such ecosystems thanks to the global and rapid accessibility of CPS services in the Cloud. However, a major challenge in this context is the highly dynamic nature of CPS services, making service composition a challenging issue. In this paper, we propose to exploit the concept of models at runtime in order to create a dynamic middleware framework that allows efficient composition of CPS services in dynamic and on-demand CPS ecosystems. This is achieved by obtaining the runtime models of CPS services and synthesizing software mediators that translate the communications between individual CPS services residing in the Cloud. The in-Cloud availability of CPS services will facilitate the design of such a middleware framework, and processing and maintaining the runtime models.