Model-Based DevOps: Foundations and Challenges

Time-to-market and continuous improvement are key success indicators to deliver for Industry 4.0 Cyber-Physical Systems (CPSs). There is thus a growing interest in adapting DevOps approaches coming from software systems to CPSs. However, CPSs are made not only of software but also of physical parts that need to be monitored at runtime. In this paper, we claim that Model-Driven Engineering can facilitate DevOps for CPSs by automatically connecting a CPS design model to its runtime monitoring, in the form of a digital twin.