Robot Brains: Circuits and Systems for Conscious Machines

Preface

Should robots understand what they are doing? What is understanding anyhow and could it be realized artificially? These questions relate to the emerging field of machine cognition and consciousness. This is a rather new discipline that is transforming philosophical musings into engineering solutions.

What is machine consciousness? Chrisley, Clowes and Torrance (2005) identify the goals of machine consciousness research as the creation of artefacts that have characteristics typically associated with awareness, self-awareness, emotion, experience, imagination, etc., and the modelling of these in embodied systems such as robots.

The first serious papers about machine consciousness began to appear in the beginning of the 1990s. Taylor (1992, 1997, 1999), Aleksander and Morton (1993) and Aleksander (1996, 2005) began to propose neural mechanisms for consciousness and mind. Duch (1994, 2005) studied a combined neural/symbolic approach to artificial minds. Trehub (1991) and Valiant (1994) presented some algorithms for cognitive processing, but did not address directly the question of machine consciousness. Sloman (2000) outlined the requirements for minds on a more general level. Philosophical approaches were presented, for example, by Dennett (1991) and Baars (1997). Also the author began his work on machine cognition in the same decade and has summarized the foundations of his approach in the book The Cognitive Approach to Conscious Machines (Imprint Academic, UK, 2003a).