Publication
Ontology-Driven Robot Design for Future Orbital and Planetary Robotics with korcut
Mehmed Yüksel; Thomas M. Röhr
European Aeronautics Science Network International Conference (EASN), 10/2022.
Abstract
The development of robots is a multidisciplinary and challenging task where the complexity grows with the capabilities and functionalities of the system, and even more when the environment imposes exceptional requirements, as in space missions. Several key points to consider when planning such a mission, e.g., the minimal space required for the system to be transported into space and its extremely compact design, multifunctionality, versatility, modularity or configurability as well as recovery purposes. Due to the interdisciplinary nature of robotic system design, robot development requires know-how of mechanical, electrical and systems engineering as well as computer science and artificial intelligence (AI). Despite the existence of various domain-specific methodologies and tools to support robot development, each tool still requires expert knowledge, and thus domain-specific experts. Ontology-driven knowledge representation for robotic system design as a method is a suitable way to collect and represent this needed knowledge about all phases of the life cycles of robotic systems in space is essential. The Q-Rock development cycle intends to overcome developmental difficulties of custom-designed robotic systems by integrating existing AI technologies to (a) enable the automated exploration of robot capabilities from robot hardware, (b) suggest robot designs that satisfy users’ needs, and (c) refine suggested robot designs. In this paper, we introduce Knowledge-based Open Robot voCabulary as Utility Toolkit (korcut) as a core component aiming to support system development for terrestrial and extraterrestrial environments, a reference implementation of the Q-Rock development cycle to support ontology-driven robot design. We use korcut to improve the robot design process from the requirements to the development phase for modular robots through the use of semantic component descriptions in state-of-the-art open-source 3D modelling software. Furthermore, the korcut ontology family includes various sub-ontologies developed to perform specific space-related tasks, such as the definition of a Standard Interconnect for On Orbit Servicing and Orbital Factory or astronaut-robot collaboration related to the Human-Machine Interaction task for lunar task domains. This paper describes korcut and provides an evaluation of its current applications from a methodological, knowledge representation and software tool perspectives.