An Architectural Android; towards research on humanoid robots

Oliver Schürer

Most recent work on domestic and office robotics is premised on one of two alternative conceptions: robots are envisaged as either instrumental or assistive. While instrumental robots take care of household chores and other productive tasks, assistive robots provide diverse kinds of services to the people. While both kinds of robots are effectively actors in social and spatial human relations, especially assistive robots will enter intimate and private situations. Some researchers, such as Sherry Turkle, the director of the MIT Initiative on Technology and Self, anticipate robots becoming a social problem. Clinical psychologist by training, she has analyzed machines as “relational artefacts” and fears that people form intimate nurturing bonds with evocative objects, including sensitive robots. This argument feeds off the old irrational fear that robots would one day enslave humans, but it does show the need to develop the mediated aspect of humanoid robots. Just like personal computers have evolved from sophisticated calculators to multimedia aggregates, robots will also turn into media as they get positioned amidst human communications. The first steps in this direction have already been taken, for example, by Anybots with their telepresence robot, the QB. Common to this branch of robotic technology is that it lacks the means to affect an environment by physical actions, a core requirement of social actors. The Architectural Android project seeks to develop applications for a humanoid that goes beyond mediating mental content alone, be it symbolic or iconic, and is capable of immediate and direct physical action. Instead of being limited to a current form of tele-presence afforded by video and sound interfaces, the robot will make use of nonverbal communication, including spatial positioning, body language, posture, gestures, and touch. By enabling the embodied representation of an individual person, the social robots of the future will make use of the empathy effects afforded by humanoid technology: they can be employed as guides at airports, museums, and other large venues during normal operation as well as emergencies; they can coach the people on matters of lifestyle or fitness as well as personal safety; they can function as realistic avatars in distant relationships where direct contact between persons is not an option.

The central challenge to set up this research project is to articulate the propositions for humanoid robots as media agents amidst social and spatial human relations via the combined expertise of software development and architecture theory. The specific goal is to enhance robot Romeo’s basic control system to the extent that it can become an active human representation as well as a social and spatial actor in everyday life situations.