Customer Reviews

Swarm Intelligence: Introduction and Applications (Natural Computing Series)

5 star:		(0)
4 star:		(2)
3 star:		(0)
2 star:		(0)
1 star:		(0)

 

 

This collection of academic papers maintains the high standards I expect from Springer, but leaves a disjointed feel. Referring to the table of contents, there are eight papers, divided between "Introduction" (half of the book) and "Applications". The four papers in Part 1 are nice, extensive surveys. Two of these, (3-4), are sufficiently specific to suggest relabeling to "Applications", leaving only two papers of broad interest. If I were a Springer marketer, I might wonder whether specialists doing research in the subjects of (3-8) would need the book, to complement the sources they already have. I myself came to it with an interest in optimization, and (a) found (2) to be a nice survey of ACO and PSO - but, again, a survey, so I would need to go to the referenced papers to see the implementation - but (b) discovered that "dynamic optimization" in (6) was not what's usually meant by the term.

You might perhaps regard this as another in Springer's extensive list of texts on robotics. Some of the book describes the biological underpinnings; the swarms that exist in nature. But the bulk relates to various implementations in robotics.

There is modelling and analysis of different swarm robotic systems. Where there is often custom hardware.

In a different light, one chapter looks at not robots, but purely computers; ie. the computers are without custom mechanical fittings that are typical of robots. Instead, routing protocols are conjectured, inspired by group behaviours found in some social insect societies like ants and bees. This is perhaps [at least to me] the most imaginative of the chapters. Those authors really did a marked conceptual shift from one context to another. Good for them!