Mobile Robotics: A Practical Introduction

Citation

Andrew, A.M. (2004), "Mobile Robotics: A Practical Introduction", Kybernetes, Vol. 33 No. 8, pp. 1336-1337. https://doi.org/10.1108/03684920410545342

Apart from the updating of references and Internet links this second edition differs from the first (reviewed in Kybernetes, 2000 Vol. 29 No. 4, pp. 525‐6) in having an extra chapter on the topic of Novelty Detection. (The increase in content is rather greater than indicated by the difference in numbers of pages, from 243, because the pages are larger in the new edition.)

The innovative nature of the contents of the book, in either edition, is not immediately apparent from the title nor from a quick flick through the pages. It contributes, as the notes on the back cover claim, to a cutting‐edge research topic. It is a pity that the title does not hint at this by the inclusion of some such term as “adaptive” or “viable”. It introduces a new approach to mobile robotics aimed at the achievement of animal‐like robustness and versatility.

In the new approach, sensors and actuators are integrated more closely than they would be in schemes based on traditional artificial intelligence (AI). It is acknowledged that this tighter linking of perception and action must be at the expense of complexity, but with the hope that the “intelligent bit” will arise as a result of interaction between various relatively simple processes, constituting an “emergent phenomenon” or “synergetic effect”.

In the new edition, the 12 case studies of the first are supplemented by one more, and they show ways of achieving an impressive repertoire of kinds of learned behaviour, navigation and map‐building. They are interesting not only from the robotics applications point of view but also for their comparison with similar performance in animals. They employ an intriguing range of biologically‐inspired techniques, including the self‐organising feature map (SOFM) due to Kohonen.

The new chapter on “novelty detection” stems from a PhD project and again makes ingenious use of the SOFM principle. Novelty detection is interesting from several points of view, one being that in small mobile robots the complexity of information processing and storage should be moderate and it is likely that attention to novelty detection will show ways of whittling down sensory data while preserving its important aspects. Moreover, there are tasks, of a surveillance or fault‐monitoring nature, where novelty detection is just what is needed. An example that is considered in the book is the exploration by a small robot of a network of underground pipes, probably sewers, to detect anything abnormal. The experiments have been done with a robot traversing corridors in a building but it is argued that the tasks are essentially similar.

An immediate problem is that a system cannot be trained to recognise novelty, which may take many forms, so the training has to be in recognising normality, after which anything that deviates sufficiently should be reported. In some earlier schemes the significant amounts of deviation have to be set manually, but in the scheme described this has been avoided by modelling the phenomenon of habituation in living systems. The model of habituation used is more recent than the historic one of Ashby (1959), which, however, might also be considered.

The book is an admirable introduction to its modern approach to mobile robotics and certainly gives a great deal of food for thought. At present, the concrete results are limited to navigation and simple tasks of box‐pushing and floor cleaning, and now surveillance and fault‐monitoring, but it is easy to feel that there is the potential for very much more and that we are being given a glimpse of principles for future animal‐like and humanoid robots. Like its first edition, this is an important and thought‐provoking book.