How to Catch a Robot Rat: When Biology Inspires Innovation

Mark Neal's delight in an exciting assessment of the state of robotics is almost lost in translation

November 18, 2010

Before getting into the meat of this review, I need to lay a few cards on the table. My wife is half-French, some of my best friends are French, and I therefore speak French (and some Spanish). I also travel a lot and do not "just shout louder" at foreigners. The pertinence of these remarks will become apparent later.

This book is a well-conceived piece of popular science writing that sets out to fill in some much-needed detail about the fairly pitiful - in comparison to, say, the Terminator - but nonetheless very exciting and attractive state of robotics as a scientific endeavour. It focuses on the current widespread use of biological metaphors, mechanisms and organisational principles for the development of robots and their control systems. Such techniques have borne valuable fruit in a range of applications, and the plethora of examples provides a good cross section of the techniques and systems that have been built and programmed using ideas borrowed from biology.

Jean-Arcady Meyer is very well placed to select, describe and comment upon the range of work in this area, and as a significant contributor to the field does a commendable job of striking a balance between providing detail and simplifying matters for the less technical reader. I am much less familiar with the work of Agnès Guillot, but the impressive breadth of coverage seen in How to Catch a Robot Rat appears to have emerged from the collaboration between the two authors.

The structure of the book often reflects the still-common view in robotics that it is sensible to separate the physical make-up of robots from their control systems, with which I am inclined to disagree. But at this point I can see angels beginning to dance on pinheads, so I will avoid travelling further down this particular avenue of discussion.

Pertinent examples, which are usually presented in bite-sized chunks, generally flow nicely into each other and range from mechanical engineering through chemical engineering to control-system and algorithm design, and serve to highlight the breadth of problems that must be solved to build effective robots.

The complexity of the problems involved in this task is often not appreciated, and tends to be glossed over when robotics is discussed in the pub, at science fairs or even in the bar at academic conferences.

Engineers building really useful and robust robots must solve every problem that evolution has solved, with the exception of reproduction, but in addition they must endow their creations with the ability to perform useful work. This book outlines a huge range of potential solutions to parts of these problems that researchers are working on, and will hopefully open the eyes of the reader to the breadth of the challenges involved and the potential solutions that nature offers.

In particular, the latter sections of the book contain significant detail about projects working towards the construction of robots that embody some of the most attractive aspects of the most hated of rodents and the most despised of insects. The sections that address the work inspired by rats and cockroaches are ably presented and well informed, and convey the essence of the book very effectively.

So far, all is well and good. Except that it isn't.

The authors originally wrote the book in French, and it was translated into English. Except that it wasn't. Not really. It was actually translated into a series of English words that are usually grammatically correct, and often convey the correct meaning if read carefully and very literally.

Sometimes, however, the translation is so direct as to render the meaning almost impenetrable, unless you have a pretty decent smattering of French in the back of your mind as you read. There are some beautiful examples of direct translation that have obviously been checked with the use of a large dictionary, but not with a native speaker.

My favourite example concerns the "ramifications" of trees. "Ramification" can be used as a term to describe the branching of trees, but the average native English speaker certainly will not read it that way until they have looked it up in their very thick dictionary, or more likely used a popular search engine and the infallible accuracy of the World Wide Web.

Most English speakers would also struggle to correctly interpret the term "chase plane"; the flying machine being referred to here is in fact usually called a fighter plane, which in French happens to translate as "avion de chasse".

In print, it is not possible to "just shout louder", so it seems that the translator has in fact adopted the other strategy: use the word you know in your own language with the richest accent in the target language that you can manage and hope for the best. This is a technique that I regularly employ when all else fails at the dinner table, and the politesse of my French relatives results either in quiet acceptance of my incompetence, or, after a few glasses of vin rouge, in gales of laughter and oaths about rugby.

Books written in English should not rely on a knowledge of the French language in order to be read by a native English speaker. The publishers should have taken better care of this, and the authors themselves (who are French speakers and writers) should have been able to rely on their services. They have been let down. I hope that they will take to the streets with their compatriots to protest not only about their government, but also about the translators they end up with.

The content of the book is occasionally marred to the point of intense irritation and incomprehensibility by such failings, but it is also in almost equal measure filled with interesting titbits and details about a huge range of topics.

It is in fact exactly the kind of book that the houses I grew up in were not filled with. I have ingrained memories of books with titles such as Intermediate Chemistry and The History of the Middle Ages, which, when opened, presented pages of dense text and little else. The upshot of this was that by the age of 12 I knew by heart almost the entire writings of Charles M. Schulz and can still name all of Snoopy's relatives from the Daisy Hill Puppy Farm. The causes of my avid reading of Snoopy's adventures were twofold. First, the books contained pictures and second, they were stored on the windowsill in the lavatory. The toilet provided an ideal retreat when there was washing-up to be done, and Snoopy provided entertainment while waiting for the clanking of crockery to stop.

I would have been delighted to find a book with a title such as How to Catch a Robot Rat that was filled with bite-sized snippets about exciting emerging technology among the Snoopy books, and I will certainly be leaving my review copy on the windowsill in the lavatory in the hope that my own children will make better use of their time in the smallest room.

Oh, and you don't need to worry about the washing-up - we have a dishwasher.

THE AUTHORS

While studying for their master's degrees in animal psychology at the University of Strasbourg in 1968, Agnès Guillot and Jean-Arcady Meyer met in the ethology lab. Guillot went on to pursue a PhD in biological science from the University of Lyon I and then a PhD in biomathematics from the University of Paris 7. She is now assistant professor in psychophysiology at the University of Paris Ouest Nanterre.

Before 1968, Meyer gained a diploma in engineering and an MSc in physics from the National Superior School of Chemistry then pursued a PhD in natural sciences at the University of Montpellier. Now an emeritus research director at the National Centre for Scientific Research, he does his research with Guillot at the Institute of Intelligent Systems and Robotics, Pierre and Marie Curie University.

Outside work, Meyer enjoys running and studying ancient Greek and Latin literature. Guillot, who spent a sabbatical interpreting songs she performed as part of "le petit Conservatoire de la Chanson", professes to now sing only in the bath, preferring to watch old Sherlock Holmes films.

Chloe Darracott-Cankovic.

How to Catch a Robot Rat: When Biology Inspires Innovation

By Agnès Guillot and Jean-Arcady Meyer

MIT Press, 232pp, £22.95

ISBN 9780262014526

Published 29 October 2010

You've reached your article limit.

Register to continue

Registration is free and only takes a moment. Once registered you can read a total of 3 articles each month, plus:

  • Sign up for the editor's highlights
  • Receive World University Rankings news first
  • Get job alerts, shortlist jobs and save job searches
  • Participate in reader discussions and post comments
Register

Have your say

Log in or register to post comments