Building artificial personalities: expressive communication channels based on an interlingua for a human-robot dance

The development of artificial personalities requires that we develop a further understanding of how personality is com-municated. This can be done through developing human-robot interaction (HRI). In this paper we report on the de-velopment of the SpiderCrab robot. This uses an interlingua based on Laban Movement Analysis (LMA) to intermediate a human-robot dance. Specifically, we developed measure-ments to analyse data in real time from a simple vision system and implemented a simple stochastic dancing algorithm on a custom built robot. This shows how, through some simple rules, a personality can emerge by biasing random behaviour. The system was tested with professional dancers and mem-bers of the public and the results (formal and anecdotal) are presented herein