Data-Driven Workflow Execution in Service Oriented IoT Architectures

System of systems within the cyber-physical context often have a greater goal to achieve - such as a product to deliver, or a series of services to be executed. Once this greater goal is described, physical elements should execute steps leading towards that goal. They often supposed to take into account the dynamically changing environment, as well as the availability and the capabilities of the surrounding systems and their services. While the Arrowhead Framework already provides support for tackling interoperability, integrability, service discovery, security and other challenges, it does not intrinsically support the execution of such, runtime workflows. The current paper proposes a supporting system - namely: the Choreographer - to be utilized for service chaining, and workflow execution. Besides describing the main services provided by the Choreographer, the principles behind its operation we detail how its control sequences fit into the orchestration and event handling procedures of Arrowhead. Furthermore, we describe a novel, Petri Net based method to demonstrate how data-driven methodology can support parallelized service execution.