A Framework for Self-adaptive Workflows in Cyber-physical Systems

Abstract

Workflows can be a useful means to formalize and enact processes among the sensors, actuators, smart objects and humans in Cyber-physical Systems (CPS). However, the dynamic nature of CPS and their resource constraint entities require the workflows and Workflow Management Systems (WfMSes) to be resilient and self-adaptive to deal with unanticipated situations and exceptions. We propose a generic framework based on the MAPE-K feedback loop to add self-management capabilities to WfMSes in the context of CPS. A case study in the smart home domain shows the general applicability of our framework for workflows and various WfMSes.