Understanding Activities and Intentions for Human-Robot Interaction

Abstract

As robots move from the factory and into the daily lives of men, women, and children around the world, it is becoming increasingly clear that the skills they will require are vastly different from the majority of skills with which they were programmed in the 20th century. In fact, it would appear that many of these skills will center on the challenge of interacting with humans, rather than with machine parts or other robots. To this end, modern-day roboticists are actively studying the problem of human-robot interaction – how best to create robots that can interact with humans, usually in a social setting. Among the many problems of human robot interaction, one of the most interesting is the problem of intent recognition: the problem of predicting the intentions of a person, usually just by observing that person. If we understand intentions to be non-observable goal-directed mental activities, then we may (quite understandably) view the intent recognition problem for robots as one of reading peoples’ minds. As grandiose as this claim may sound, we believe that this understanding of intent recognition is quite reasonable; it is this interpretation that we seek to justify in the following pages. Every day, humans observe one another and on the basis of their observations “read people’s minds,” correctly inferring the intentions of others. Moreover, this ability is regarded not as remarkable, but as entirely ordinary and effortless. If we hope to build robots that are similarly capable of successfully interacting with people in a social setting, we must endow our robots with an ability to understand humans' intentions. In this paper, we review the intent recognition problem, and provide as an example a system we have been developing to recognize human intentions. Our approach is ultimately based on psychological and neuroscientific evidence for a theory of mind (Premack & Woodruff, 1978), which suggests that the ease with which humans recognize the intentions of others is the result of an innate mechanism for representing, interpreting, and predicting other's actions. The mechanism relies on taking the perspective of others (Gopnick & Moore, 1994), which allows humans to correctly infer intentions. Although this process is innate to humans, it does not take place in a vacuum. Intuitively, it would seem that our understanding of others' intentions depend heavily on the contexts in which we find ourselves and those we observe. This intuition is supported by Source: Human-Robot Interaction, Book edited by: Daisuke Chugo, ISBN 978-953-307-051-3, pp. 288, February 2010, INTECH, Croatia, downloaded from SCIYO.COM