The use of voice control for industrial robots in noisy manufacturing environments

In the present study, the efficiency of voice control for industrial robot systems is analyzed using a typical manufacturing environment where a speech interaction system reveals strong advantages. Challenges industrial environments bear for speech recognition systems are mainly found in the high-volume and dynamically changing background noises. The effects of such noises are analyzed in detail using a joinery as showcase. As a robust solution for very noisy environments the use of a throat microphone is proposed. Due to their special sensing technique, these microphones present excellent noise cancelling characteristics while the frequency content of the speech signal differs significantly from the sounds recorded using a standard microphone. Two approaches for the integration of a throat microphone into an existing speech application are discussed: using the throat microphone a) as an additional sensor or b) as an alternative to a standard headset microphone.