Semantic Inferences Towards Smart IoT-Based Systems Actuation Conflicts Management

IoT-based systems have long been limited to collecting field information via sensors distributed at the edge of their infrastructure. However, in many areas such as smart home, smart factory, etc. these systems include devices that interact with the physical environment via common actuators. Throughout the lifecycle of these systems, from design, to deployment to operation, the ability to avoid actuation conflicts, both in terms of the commands that actuators receive (direct conflicts) and the effects that they produce (indirect conflicts), is a new challenge in the realm of trustworthy Smart IoT-based Systems (SIS). As part of the European project ENACT, which aims to provide full DevOps support for trustworthy SIS, we present a lightweight ontology that provides SIS designers with (1) a semantic metamodel to formally describe SIS subsystems and the actuators they interact with, and (2) a set of SWRL (Semantic Web Rule Language) inference rules to automatically identify and semi-automatically resolve actuation conflicts. Consistent with the best practices of the DevOps approach, a particular emphasis is placed on facilitating the use and interpretation of inference results. To provide insight into the appropriateness of the proposed approach in the context of SIS, rule processing times for different actuation conflict configurations are provided.