When industrial robots become social: On the design and evaluation of a multimodal interface for welding robots

Abstract

Lately, technological advances made possible the development of complex robotic systems in many industry areas. However, such systems require skilled operators undergoing long hours of intensive training. In this paper, we present the design and evaluation of an intuitive multimodal user interface that connects to an intelligent robot welder. The interface is meant to enable less trained workers to interact with the robot in a mo Workers are able to validate welding settings, ask the robot for advice and correct errors before starting the welding process using a multimodal approach, i.e. using speech, type or mouse/finger touch. To evaluate our interface, we performed two consecutive studies: a qualitative evaluation involving 4 participants and a quantitative evaluation with 15 participants. Results showed that our multimodal concept brings added value for users interacting with the robot welder while the interface is pleasant and task-supportive. We received several suggestions for improvement that will be incorporated in the next development iteration.