Speech Interface for a Lightweight Industrial Robot

This paper presents an interface used to control a lightweight industrial robot via speech. Many examples of speech control can be found in everyday life, and there may also be a useful application in industrial automation devices, e.g., robots. Because of their intuitive teach-in concept for hand guidance, lightweight robots are particularly convenient. Furthermore, they include special safety functions in the form of speed and force limitations, which makes the combination of speech recognition and industrial robots in the first instance uncritically. In this study, speech control was firstly compared with other approaches to intuitive and novel robot programming. Here, it is important to distinguish between experts and non-experts. A suitable speech-recognition system was chosen for implementation of a robot speech-control interface. For this purpose, we compared various aspects of currently available speech software. Speech-recognition functionality was then fitted to the robot, and program flow-charts for speech control were developed. Speechcontrolled teach-in and speech-controlled robot motion were both integrated into the robot system, and the algorithms were successfully verified by practical experiments. Finally, speech teach-in was compared to teach-in using the teach pendant, and to teach-in of a conventional industrial robot, in terms of duration and usability.