Towards Self-Configuring Plug & Produce Robot Systems Based on Ontologies

Abstract

The industry is undergoing a disruptive shift from mass production to individually manufactured one-offs. To achieve this transformation, the production facilities must also be flexible and able to configure themselves according to requirements. Often it is not enough to reconfigure the software, but the hardware must also be adapted accordingly. Such customizable hardware and software systems are called Plug & Produce systems. For such Plug & Produce robot systems, we have developed an automatic configuration generation for hardware configurations based on capability/skill descriptions of individual hardware components. The configuration creation is done with the help of an ontology, in which information about modular hardware components, their capabilities, and the interrelationships of their interfaces (geometric, electrical, and data) is stored.