Abstract/Summary

Cyber-Physical Systems (CPSs) are systems of systems merging the physical world with the virtual world of cyberspace. CPSs are often highly complex and interdisciplinary. Thus, their design, analysis, development, performance refinement, and testing require the consideration of multiple aspects from various domains. This is even more critical for Machine Learning (ML) enabled CPSs, which are increasingly becoming more prevalent. In this study, we take the first step towards collecting a set of quality attributes for smart CPSs. Further, we introduce a set of merit criteria for tools that can be used for designing, modeling, and developing smart CPSs. This is to provide a framework for the comparison and benchmarking of smart CPSs and CPS modeling tools. The framework has been put through the first phase of expert interview-based validation.