Model-based documentation of dynamicity constraints for collaborative cyber-physical system architectures: Findings from an industrial case study

Abstract

Collaborative cyber-physical systems form highly dynamic networks, consisting of collaborating individual systems at run-time. These networks provide functionality that goes beyond the functionality of the individual systems that are part of the network. The availability of such collaborative functionalities depends not only on the individual systems’ behaviors but also on the current morphology of the network (i.e., the architecture of the overall network that defines what kind of systems and how many systems are actually partaking in the collaboration at that moment). It is of importance to thoroughly consider the possible morphologies of the network of cyber-physical systems during the development of the individual systems as these are sources for requirements as well as design decisions. These requirements and design decisions must be associated to the particular morphologies they are applicable to in order to ensure traceability and comprehensibility between the morphology of the network of cyber-physical systems and the individual systems. Hence, we aim at defining an orthogonal model which allows to document dynamicity constraints and therefore restrain the potential morphologies a network can take. This paper contributes results from an exploratory case study conducted in the industry automation domain. In the case study, an existing variability modeling approach was applied to the development of a network of autonomous transport robots. Based on the variability modeling approach an orthogonal dynamicity constrained model was defined and evolved to fit industry's needs. Results show that the use of an orthogonal model for dynamicity constraints can improve the development of cyber-physical systems designed to participate in collaborative networks.