Situation-Aware Execution and Dynamic Adaptation of Traditional Workflow Models.

The continuous growth of the Internet of Things together with the complexity of modern information systems results in several challenges for modeling, provisioning, executing, and maintaining systems that are capable of adapting themselves to changing situations in dynamic environments. The properties of the workflow technology, such as its recovery features, makes this technology suitable to be leveraged in such environments. However, the realization of situation-aware mechanisms that dynamically adapt process executions to changing situations is not trivial and error prone, since workflow modelers cannot reflect all possibly occurring situations in complex environments in their work-flow models. In this paper, we present a method and concepts to enable modelers to create traditional, situation-independent workflow models that are automatically transformed into situation-aware workflow models that cope with dynamic contextual situations. Our work builds upon the usage of workflow fragments, which are dynamically selected during runtime to cope with prevailing situations retrieved from low-level context sensor data. We validate the practical feasibility of our work by a prototypical implementation of a Situation-aware Workflow Management System (SaWMS) that supports the presented concepts.