Can knowledge-based offline-programming compensate for the lack of welding specialists?

Automation with robots is an obvious opportunity to cope with the lack of qualified welders, especially in small and medium-sized companies. The key requirement to achieve the same flexibility and economic efficiency as with human welders is a quick and easy programming of the robots even for smallest batch sizes. This contribution discusses the preconditions of new methods for fast robot programming, describes solutions for their implementation and illustrates their realization using the example of the offline-programming system MOSES-AI. This system is already capable of autonomously programming welding robots with a degree of automation of up to 100% for a number of applications including automatic welding process planning, automatic motion planning and simulation-based program generation.