Compositional Engineering Frameworks for Development of Smart Cyber-Physical Systems: A Critical Survey of the Current State of Progression

Abstract

Various system-engineering frameworks (SEFs) have been developed for composable systems, whose overall operation is the sum of the operations of their components. However, smart cyber-physical systems (S-CPSs) are compositional in nature. Their reasoning capability and system knowledge assume an ‘ampliative’ (inter)operation of all hardware, software, or cyberware components. The need for SEFs that support synthesis, modeling, analysis, simulation, verification, and validation of S-CPSs is recognized in the literature. The objective of this paper is a critically review the state of development of compositionality enabling frameworks. Both quantitative and qualitative literature studies were conducted in combination with critical system thinking. The reasoning model used in the qualitative analysis was derived based on the findings of the quantitative analysis. The major observations are: (i) the notion of compositionality is not studied extensively in the context of S-CPSs that do not obey the principle of reductionism, (ii) methodological support of implementation of compositional CPSs seems to be in its infancy, (iii) SEFs may play a crucial role in synthesis, modeling and implementation of S-CPSs, and (iv) SEFs for compositional system design may be realized using the principles of semantic knowledge fusion or meta-synthesis. Our follow up research targets a formal definition and computational implementation of a testable prototype of a specific SEF tool supporting compositional design of reasoning mechanisms.