Embodied Evolution: Distributing an evolutionary algorithm in a population of robots

Abstract

We introduce Embodied Evolution (EE) as a new methodology for evolutionary robotics (ER). EE uses a population of physical robots that autonomously reproduce with one another while situated in their task environment. This constitutes a fully distributed evolutionary algorithm embodied in physical robots. Several issues identified by researchers in the evolutionary robotics community as problematic for the development of ER are alleviated by the use of a large number of robots being evaluated in parallel. Particularly, EE avoids the pitfalls of the simulate-and-transfer method and allows the speed-up of evaluation time by utilizing parallelism. The more novel features of EE are that the evolutionary algorithm is entirely decentralized, which makes it inherently scalable to large numbers of robots, and that it uses many robots in a shared task environment, which makes it an interesting platform for future work in collective robotics and Artificial Life. We have built a population of eight robots and successfully implemented the first example of Embodied Evolution by designing a fully decentralized, asynchronous evolutionary algorithm. Controllers evolved by EE outperform a hand-designed controller in a simple application. We introduce our approach and its motivations, detail our implementation and initial results, and discuss the advantages and limitations of EE.

Comparison of Selection Methods in On-Line Distributed Evolutionary Robotics

“…PGTA (Probabilistic Gene Transfer Algorithm) introduced by [15], is commonly cited as the first implementation of a distributed on-line ER algorithm. This algorithm evolves the weights of fixed-topology neural controllers and agents exchange parts (genes) of their respective chromosomes using local broadcasts.…”

“…On the other hand, on-line evolution ( [15]) takes a different approach in which behavior learning is performed during the actual execution of a task. In these algorithms, learning or optimization is a continuous process, i.e.…”

Embedded Evolutionary Robotics: The (1+1)-Restart-Online Adaptation Algorithm

“…Embedded is defined as both online (the adaptation/learning process never stops) and onboard (the optimization algorithm and evaluation process are part of the control loop). To date, only few, but promising, works have adressed this topic: [7,11,12,14,15,17,20,23,24,25,26]. Running an embedded EA within a single robot provides strong advantages regarding continuous adaptation and autonomy with regards to a human supervisor.…”

It’s Fate: A Self-organising Evolutionary Algorithm

“…This provides a natural solution to the problems of the limited range and scalability a single oracle based EA would suffer from. Compared to alternative solutions where evolutionary operators are embodied in the robots [13,7] Fate Agents offer increased controllability for experimenters and users. Furthermore, our Fate Agents are not only operating on candidate solutions, but also on themselves.…”