Evolutionary Robotics and Open-Ended Design Automation - Biomimetics: Biologically Inspired Technologies - page 148

Figure 4.9 (See color insert following page 302) (a) Reconfigurable molecube robots. (From Zykov, V.,
Mytilinaios, E., Lipson, H., (2005) Nature, 435 (7038), 163-164. With permission.) (b) Stochastic modular robots
reconfigure by exploiting Brownian motion, and may allow reconfiguration at a micro-scale in the future. (From
White, P. J., Kopanski, K., Lipson, H. (2004) Stochastic self-reconfigurable cellular robotics, IEEE International
Conference on Robotics and Automation (ICRA04). With permission.) (c) Rapid prototyping. (d) Future rapid
prototyping systems will allow deposition of multiple integrated materials, such as elastomers, conductive wires,
batteries, and actuators, offering evolution of a larger design space of integrated structures, actuators, and
sensors, not unlike biological tissue. (From Malone, E., Lipson, H. (2004) Functional freeform fabrication for
physical artificial life, Ninth International Conference on Artificial Life (ALIFE IX), Proceedings of the Ninth
International Conference on Artificial Life (ALIFE IX). With permission.)

Little formal design knowledge . Evolutionary algorithms are ''knowledge sparse''; they essentially
generate knowledge through search. They are thus able to work in the absence of formal knowledge
in the problem domain. Given enough time and resources, one may be able to design a specialized
algorithm that takes advantage of specific domain knowledge and outperforms an evolutionary
algorithm, but often this is time consuming, costly, and too difficult.

.