Automated Derivation of Application Workload Models for Design Space Exploration of Industrial Distributed Cyber-Physical Systems

Manufacturing companies of complex distributed cyber-physical systems (dCPS) are encountering challenges with respect to designing their next-generation machines. They need efficient Design Space Exploration (DSE) techniques to evaluate possible design decisions and their consequences on nonfunctional aspects of the systems. To enable scalable and efficient DSE of complex dCPS, it is essential to have abstract and coarse-grained models that are both accurate and capable of capturing dynamic application workloads. This paper addresses the scientific challenge of defining and automatically deriving an application workload model for DSE of complex dCPS. Our approach leverages trace analysis to derive a dynamic workload model that accurately represents computation and communication actions within an application in a timing agnostic manner. We demonstrate the effectiveness of our approach by applying it to the ASML Twinscan lithography machine, which is a real-world complex dCPS. The results of our study demonstrate an accuracy of over 90% in capturing the real application workloads.