Evolving controllers for a homogeneous system of physical robots: structured cooperation with minimal sensors

Abstract

We report on recent work in which we employed artificial evolution to design neural network controllers for small, homogeneous teams of mobile autonomous robots. The robots were evolved to perform a formation-movement task from random starting positions, equipped only with infrared sensors. The dual constraints of homogeneity and minimal sensors make this a non-trivial task. We describe the behaviour of a successful system in which robots adopt and maintain functionally distinct roles in order to achieve the task. We believe this to be the first example of the use of artificial evolution to design coordinated, cooperative behaviour for real robots.

Evolution of Implicit and Explicit Communication in Mobile Robots

“…The theoretical approach and the methodology followed in this chapter is in line with the first Chapter of this section and with the work of Baldassarre et al (2003);Di Paolo (1997; Marocco and Nolfi (2007);Quinn (2001) ;Quinn et al (2003); Trianni and Dorigo (2006). However, the experimental scenario proposed here is more advanced than in the experimental works mentioned above with respect to the following aspects (or with respect to the possibility to study the following aspects in combination): (i) the complexity of the chosen task that allows us to study how several behavioural and communication skills are developed and co-adapted during the evolutionary process, (ii) the richness of the agents' sensory-motor system that supports, for example, the exploitation of both explicit and implicit communication, (iii) the validation of the results obtained in simulation in hardware.…”

Selection methods regulate evolution of cooperation in digital evolution

“…While all five selection methods are frequently used to simulate differential selection (PSM in [19][20][21][22][23][24][25][26][27][28][29][30][31][32]; RSM in [33,34]; TPSM in [35][36][37]; TUSM in [38][39][40][41][42][43][44][45][46], TSM in [22,47,48]), the choice between them is rarely justified. Moreover, little attempt has been made to quantify the effects of selection methods on the dynamics of the digital evolution (but see [22,49]).…”

“…To address this problem, we investigate theoretically and with numerical experiments how the five selection methods regulate the evolution of cooperation. We focus on cooperation, because digital evolution is especially popular in this domain [19][20][21][22][23][24][26][27][28][29]33,38,41,47,48,54,55], and it is an important biological phenomenon that has attracted extensive scientific interest (see [56][57][58][59][60] for reviews). We consider a population of related individuals, each having a genotype that consists of a haploid allele encoding for cooperation or defection.…”

Evolution of Signalling in a Group of Robots Controlled by Dynamic Neural Networks

“…Obviously this work does not provide enough evidence to answer positively. Any communication system that escapes from the local and simple interactions (e.g., communication through infra-red sensors-see [6]) might present disadvantages as well as advantages. In fact, when we move from a robot-to-robot to a robot-to-many interaction, not only the benefit of the knowledge of the environment acquired, but also possible errors spread faster.…”

“…Based on artificial evolution, ER finds sets of parameters for artificial neural networks (ANN's) that guide the robots to the accomplishment of their objective. ER can be employed to look at the effects that the physical interactions among embodied agents and their world have on the evolution of individual behaviour and social skills (see [6]). ER also permits the co-evolution of communicative and noncommunicative behaviour, since it lets different characteristics co-adapt, only requiring an overall evaluation of the group (see [7]).…”