Augmented Reality Supported Modeling of Industrial Systems to Infer Software Configuration
Matúš Tomlein, Kaj Grønbæk · Proceedings of the ACM on Human-Computer Interaction · 2018
This paper proposes and evaluates an approach for building models of installed industrial Cyber-Physical Systems using augmented reality on smartphones. It proposes a visual language for annotating devices, containers, flows of liquids and networking connections in augmented reality. Compared to related work, it provides a more lightweight and flexible approach for building 3D models of industrial systems. The models are further used to automatically infer software configuration of controllable industrial products. This addresses a common problem of error-prone and time-consuming configuration of industrial systems in the current practice. The proposed approach is evaluated in a study with 16 domain experts. The study participants are involved in creating a model of an industrial system for water treatment. Their comments show that the approach can enable a less error-prone configuration for more complex systems. Opportunities for improvement in usability and reflections on the potential of the approach are discussed.