Non-Verbal Auditory Aspects of Human-Service Robot Interaction

As service robots become ever more pervasive, the number, degree and depth of interaction with humans, particularly fellow workers, is increasing rapidly. Humans are generally shaped alike, respond in predominantly similar ways and are often inherently predictable to other humans. Robots, by contrast, have an exceptional diversity of size, shape, mobility, function, and their intentions or actions are often less predictable. Humans working in close proximity have learnt to provide cues to their behaviour, both verbal and non-verbal, and we argue that this is an important aspect of maintaining both safety and comfort in a mixed work or social environment. At present, robots do not provide any such cues to their fellow workers, which can be cause of human discomfort, and indeed contribute to safety issues when working in close proximity to humans. This paper considers the non-verbal auditory aspects of interaction in a work environment, with particular emphasis on safe and comfortable integration of service robots into such locations. In particular, we propose a classification of interaction levels to inform the construction, programming and operation of robots in the workplace.