Multi-layer IIoT Architecture for Autonomous Vehicles: A Proof of Concept

Abstract

Autonomous driving relies on various sensors and a robust communication between the vehicle and interacting IT nodes. This concept requires both, utilizing data for track-specific models as well as data analytics and feedback control during driving. While modeling is traditionally an offline task, the analysis and control of the vehicle has to be managed online. These two modes differ considerably regarding hardware and software requirements, affecting the design of the Industrial Internet of Things (IIoT) architecture. Thus, for the application of autonomous vehicles a multi-layer IIoT architecture is proposed addressing the challenges from the field level to the cloud. The development introduced in this paper was conducted in an industrial research project by a student group working in the research and learning factory, the Smart Production Lab of the FH JOANNEUM in Kapfenberg, Austria. In a proof of concept, the IIoT architecture has been validated using the IIoT infrastructure of the learning factory. Two main results could be achieved. First, the technical perspective shows that the proposed architecture provides a robust IIoT framework for the specific use case in the laboratory environment. Second, the students acquired a deep understanding of the concepts and gained experience in the interaction with industrial clients.