Modelling complex and flexible processes for smart cyber-physical environments

•Analysis of challenges for process modelling in cyber-physical systems (dynamics, heterogeneity, complexity, parallelism, evolution, distribution).•Proposition of a component-based, object-oriented meta-model for processes in cyber-physical system.•Discussion of semantic domain-specific annotations for runtime adaptation.•Evaluation of polymorphism and semantic information for dynamic and flexible process execution.•Presentation of a model-based process modelling toolchain.

Cyber-physical systems introduce new requirements for modelling and executing autonomous processes. Current workflow languages are not able to completely fulfil these requirements, as they lack expressiveness and flexibility. In this paper, we present an object-oriented workflow language for formalizing processes within heterogeneous and dynamic environments. Its component-based meta-model enables the hierarchical composition of processes and process variants. Domain-specific typing and polymorphism leverage the flexibility of processes by enabling the dynamic selection of process components. We present a modelling environment and a distributed execution engine for the meta-model. In addition, we discuss the use of semantic technologies for smart workflows.