Augmented Reality Training System

Ongoing development of automation technologies and a foreseeable lack of specialized workforces require a better organization of work processes in production and adjacent domains. In this context, an optimized division of functions and labour between humans and robots in terms of direct human-robot cooperation will become a relevant topic. But from a security and labour protection point of view, this is critical, since the strict separation between the robot’s danger area and the employee has to be given up. The earlier and the more effective the planned process can be checked for compliance with security directives, the higher are security and quality levels as well as cost and time savings of the installation. Today’s 3D simulation environments cover industrial robotics comprehensively already: from heavyweight goods handling to arc welding and paint shop applications, diverse scenarios can be modelled and simulated. If necessary, human workers are represented by digital human models (Ramsis, Jack AnyBody etc.). Because of their complicated handling and their high number of degrees of freedom, these models are most often not adequate for interactive applications, especially when dealing with security-critical interaction between humans and robots. Deployment of an Augmented Reality (short: AR) system where the user himself takes over the egocentric view of the worker to check and, if necessary, to correct the whole process is more self-suggestive here. This is why a flexible AR system for simulation and validation of direct human-robot cooperation in industrial use cases has been conceptualized and developed. With it, cooperation processes shall be inspected for security lacks more comprehensively than ever. By combining several different approaches and technologies, of which some have been rather untypical for industrial applications, as far as now, a maximum in realism shall be achieved to significantly increase the user’s immersion in the virtual environment.