Evolutionary robotics applies the selection, variation, and heredity principles of natural evolution to design robots with embodied intelligence. It can be considered as a subfield that aims create more robust adaptive robots. A pivotal feature evolutionary approach is it considers whole robot at once, enables exploitation features in holistic manner. also seen an innovative study based on new kind experimentalism. The use substrate help address questions are difficult, if not impossible,...

This paper investigates the role of communication in improving coordination within robot swarms, focusing on a paradigm where learning and execution occur simultaneously decentralized manner. We highlight can play addressing credit assignment problem (individual contribution to overall performance), how it be influenced by it. propose taxonomy existing future works communication, information selection physical abstraction as principal axes for classification: from low-level lossless...

This article is concerned with a fixed-size population of autonomous agents facing unknown, possibly changing, environments. The motivation to design an embodied evolutionary algorithm that can cope the implicit fitness function hidden in environment so as provide adaptation long run at level population. proposed algorithm, termed mEDEA, shown be both efficient unknown environments and robust abrupt unpredicted changes environment. emergence consensus towards specific behavioural strategies...

Evolutionary robotics is heading towards fully embodied evolution in real-time and real-space. In this paper we introduce the Triangle of Life, a generic conceptual framework for such systems which robots can actually reproduce. This be instantiated with different hardware approaches reproduction mechanisms, but all cases system revolves around conception new robot organism. The other components capture principal stages system; as whole serves guide realizing anticipated breakthrough...

HAL is a multi-disciplinary open access archive for the deposit and dissemination of scientific research documents, whether they are published or not.The documents may come from teaching institutions in France abroad, public private centers.L'archive ouverte pluridisciplinaire HAL, est destinée au dépôt et à la diffusion de scientifiques niveau recherche, publiés ou non, émanant des établissements d'enseignement recherche français étrangers, laboratoires publics privés. The emergence...

Whereas naturally occurring swarms thrive when crowded, physical interactions in robotic are either avoided or carefully controlled, thus limiting their operational density. Here, we present a mechanical design rule that allows robots to act collision-dominated environment. We introduce Morphobots, swarm platform developed implement embodied computation through morpho-functional design. By engineering three-dimensional printed exoskeleton, encode reorientation response an external body force...

Embodied evolutionary robotics is a sub-field of that employs algorithms on the robotic hardware itself, during operational period, i.e., in an on-line fashion. This enables systems continuously adapt, and are therefore capable (re-)adjusting themselves to previously unknown or dynamically changing conditions autonomously, without human oversight. paper addresses one major challenges such face, viz. robots must satisfy two sets requirements. Firstly, they continue operate reliably their...

In this paper, we present an implementation of social learning for swarm robotics. We consider as a distributed online reinforcement method applied to collective robots where sensing, acting and coordination are performed on local basis. While some issues specific artificial systems, such the general objective efficient (and ideally, optimal) behavioural strategies fulfill task defined by supervisor, other shared with in natural systems. discuss these issues, paving way towards cumulative...

Roborobo! is a multi-platform, highly portable, robot simulator for large-scale collective robotics experiments. coded in C++, and follows the KISS guideline ("Keep it simple"). Therefore, its external dependency solely limited to widely available SDL library fast 2D Graphics. based on Khepera/ePuck model. It targeted single multi-robots simulation, has already been used more than dozen published research mainly concerned with evolutionary swarm robotics, including environment-driven...

Biomimetic robots are promising tools in animal behavioural studies. If they socially integrated a group of animals, can produce calibrated social stimuli to test the responses. However, design such is challenging as it involves both luring capability including appropriate robot behaviours, and acceptation by animals companions. Here, we investigate integration biomimetic driven models into zebrafish (Danio rerio). The behaviours based on stochastic model linking visual perception individual...

Social interactions involving coordination between individuals are subject to an "evolutionary trap." Once a suboptimal strategy has evolved, mutants playing alternative counterselected because they fail coordinate with the majority. This creates detrimental situation from which evolution cannot escape, preventing of efficient collective behaviors. Here, we study this problem using evolutionary robotics simulations. We first confirm existence trap in simple setting. then, however, reveal...

Whereas naturally occurring swarms thrive when crowded, physical interactions in robotic are either avoided or carefully controlled, thus limiting their operational density. Here we present a mechanical design rule that allows robots to act collision-dominated environment. We introduce the Morphobots -- swarm platform developed implement embodied computation through morpho-functional design. By engineering 3D-printed exoskeleton encode re-orientation response an external body force (such as...

We extend the study of inertial effects on two-dimensional dynamics active agents to case where self-alignment is present. In contrast with most common models particles, we find that self-alignment, which couples rotational translational one, produces unexpected and nontrivial dynamics, already at deterministic level. Examining first motion a free particle, role inertia depending sign self-aligning torque. When positive, does not alter steady-state linear an a-chiral self-propelled particle....

Robot behavior learning by demonstration deals with the ability for a robot to learn from one or several demonstrations provided human teacher, possibly through tele-operation imitation. This implies controllers that can address both (1) feature selection problem related great amount of mostly irrelevant sensory data and (2) dealing temporal sequences demonstrations. Echo state networks (H. Jaeger, 2001) have been proposed recently time series prediction shown perform remarkably well on this...