ABSTRACT
The fascination of collective and swarm systems is huge: social
insects, fish, and animals can successfully solve various prob-
lems [Bonabeau et al. (1999)]. By applying the principles of swarms to robotics, we can mimic several capabilities of natural collective
systems [Sahin (2004)], in particular their amazing ability towork at
different scales, including the number of collective agents, diversity,
cooperativity of behavior, load, and several other parameters.
Generally, these so-called “scalability metrics” are related to the
notion of scalability [Constantinescu et al. (2004)]. Scalability is a key feature of collective systems because it
integrates many other issues, such as reliability, self-repairing, and
several economic factors. Especially important is scalability in the
micro-and molecular domains, which traditionally deal with a huge
number of components [Balzani et al. (2003)]. It is also assumed that by making collective systems larger and more complex, we
can achieve more functional capabilities and, finally, more extended
“collective intelligence.” In this chapter, we will investigate the
dependencies between scalability metrics and the informational,
structural, functional, and behavioral aspects of collective systems.