Open-ended evolution as a means to self-organize heterogeneous multi-robot systems in real time

Distributed embodied evolution for collective tasks

“…The canonical distributed EE used here is based in the AsiCo (Asynchronous Situated Coevolution) algorithm developed by Prieto et al [3]. This algorithm has been reduced to its "barebones", isolating the processes that guide individual evolution from the specific scenario.…”

“…ACM 978-1-4503-1964-5/13/07. Genotypic recombination: it is performed as a combination of two recombination strategies, a local search operator and a bipolar crossover (see [3] for details). It has been defined here as the probability of using the local search (P ls ).…”

Behavioral Specialization in Embodied Evolutionary Robotics: Why So Difficult?

“…In recent years, the on-line nature of such algorithms was shown to be very robust when conducting experiments with real robots (Watson et al, 2002;Prieto et al, 2010;Bredeche et al, 2012;Trueba et al, 2013): compared with more classic evolutionary robotics setup, the emphasis in EER is on the design of robust algorithms (i.e., design while already deployed) rather than on producing robust solutions (i.e., design then deploy) (Doncieux et al, 2015;Silva et al, 2016). However, the complexity of the tasks to be achieved has been quite limited so far, either resulting with each individual maximizing its own benefit [e.g., phototaxis (Watson et al, 2002), foraging, exploration, etc.]…”

Applying the canonical distributed Embodied Evolution algorithm in a collective indoor navigation task

“…Consequently, the problem that appears is how to evaluate on line the candidate genotypes for each embodied individual [8]. The second variation, distributed Embodied Evolution (dEE), follows the same scheme as the original algorithm proposed in [7] where each individual carries only its own genotype as it occurs in natural evolution, and mating occurs in a completely decentralized fashion [1] [13]. Due to its higher level of biological fidelity and the complexity of the dynamics which derive from simple interactions between individuals, this research is focused in the study of the dEE variation.…”

“…The suitability of EE for multi-robot tasks has led to the development of different algorithms, but basically three of them must be highlighted in the dEE approach: mEDEA [2] [3], PGTA [7] [21] and ASiCo [13] [19]. They have been applied with success to different tasks requiring self-organization, adaptation, emergence of specialization, and so on.…”