A key aspect of decentralized multirobot coordination is communication. However, beyond simple signaling, there are only few reports in the literature on the successful evolution of communication, with successes largely dependent on specific tasks and evolutionary setups. Thus, there is a lack of standardized communication frameworks that can be applied to different tasks without the need to redesign, rebuild, or re-evolve the entire system for every new task. In this article, we propose a novel communication module that does not need to be modified for its use in different tasks. Each robot has a coordinate (state) in a virtual communication space. The communication space is partitioned into virtual regions, and each region is linked to a physical behavior, such as seeking resources, phototaxis, or recharging the battery. A robot's individual behavior is determined by the region to which its current communication state belongs. Since robots can navigate the communication space and continually broadcast their coordinates to neighbors within range, robot swarms can effectively coordinate their behavior in a self-organized manner. We demonstrate that the same evolved communication module is effective in three swarm robotics tasks: 1) the physical aggregation of the robots into groups of a desired size; 2) the formation of desired swarm geometries; and 3) a foraging task based on temporal role allocation. The results show that the communication module provides good and scalable performance in all tasks, representing a significant step toward a task-agnostic communication framework for robot swarms.

Path planning is a vitally important ability for autonomous mobile robots. Because of the high computational complexity, the optimal solution is generally infeasible since the required computation time increases exponentially with the increase in the problem size. Instead, it is common to rely on heuristic and meta-heuristic algorithms to find near-optimal solutions. Although heuristic algorithms have proven effective and computationally inexpensive in small workspaces, they do not always scale to large environments and tend to get trapped in local minima. Also, while meta-heuristic algorithms are attracting considerable attention because of their effectiveness in optimization, they still require significant computational resources and are non-deterministic. In this paper, we introduce a novel Fast Constructive Algorithm (FCA) for deterministic path optimization that requires comparatively few computational resources to generate an optimized path. Our proposed FCA efficiently generates the waypoints for a path based only on the obstacles that intersect with the straight-line segment linking the robot's current position to the target location. The key idea is to construct the path by iteratively calculating the best waypoint to avoid the next obstacle in the robot's path. The effectiveness of the FCA is assessed on several maps with distinct complexities and its performance is compared with different state-of-the-art path planning algorithms. Our results show that the proposed FCA is competitive and can outperform existing algorithms in terms of path length and computation time.

Spiking Neuronal Networks (SNNs) realized in neuromorphic hardware lead to low-power and low-latency neuronal computing architectures. Neuromorphic computing systems are most efficient when all of perception, decision making, and motor control are seamlessly integrated into a single neuronal architecture that can be realized on the neuromorphic hardware. Many neuronal network architectures address the perception tasks, while work on neuronal motor controllers is scarce. Here, we present an improved implementation of a neuromorphic PID controller. The controller was realized on Intel's neuromorphic research chip Loihi and its performance tested on a drone, constrained to rotate on a single axis. The SNN controller is built using neuronal populations, in which a single spike carries information about sensed and control signals. Neuronal arrays perform computation on such sparse representations to calculate the proportional, derivative, and integral terms. The SNN PID controller is compared to a PID controller, implemented in software, and achieves a comparable performance, paving the way to a fully neuromorphic system in which perception, planning, and control are realized in an on-chip SNN.

The ability to reliably detect faults is essential in many real-world tasks that robot swarms have the potential to perform. Most studies on fault detection in swarm robotics have been conducted exclusively in simulation, and they have focused on a single type of fault or a specific task. In a series of previous studies, we have developed a robust fault-detection approach in which robots in a swarm learn to distinguish between normal and faulty behaviors online. In this paper, we assess the performance of our fault-detection approach on a swarm of seven physical mobile robots. We experiment with three classic swarm robotics tasks and consider several types of faults in both sensors and actuators. Experimental results show that the robots are able to reliably detect the presence of hardware faults in one another even when the swarm behavior is changed during operation. This paper is thus an important step toward making robot swarms sufficiently reliable and dependable for real-world applications.

Evolutionary robotics approaches have traditionally been focused on monolithic controllers. Recent studies on the evolution of hierarchical control have, however, yielded promising results. Hierarchical approaches typically rely on a repertoire of behaviour primitives (which themselves can be the result of an evolutionary process), and an evolved top-level arbitrator that continually executes primitives from the repertoire to solve a given task. In this paper, we compare different controller architectures for the evolution of top-level arbitrators. We propose two new methods, one based on neural networks and another based on decision trees induced by genetic programming. We compare the new approaches with existing ones, namely neural network regressors and non-hierarchical control, in a challenging simulated maze navigation task that requires a broad diversity of primitives. Based on empirical results, we draw a number of conclusions regarding the strengths and limitations of each of the studied approaches.

The evolution of task-oriented control for robots with complex locomotor systems is currently out of reach for traditional evolutionary computation techniques, as the coordination of a high number of locomotion parameters in response to the robot's sensory inputs is extremely challenging. Evolutionary techniques have therefore mainly been applied to the optimization of specific locomotion patterns, such as forward motion. In this paper, we explore the evolutionary repertoire-based control (EvoRBC) approach, which divides the synthesis of control into two steps: 1) the evolution of a repertoire of locomotion primitives using a quality diversity algorithm and 2) the evolution of a high-level arbitrator that leverages the locomotion primitives in the repertoire to solve a given task. We comprehensively study the main components of the EvoRBC approach using a four-wheel steering robot. We then conduct a set of experiments in simulation using a hexapod robot. Our results show that EvoRBC is robust to parameter variations, and for all the robots tested, it is able to evolve controllers for a maze navigation task and significantly outperforms the traditional evolutionary robotics approach.

We propose Hyb-CCEA, a cooperative coevolutionary algorithm for the evolution of genetically heterogeneous multiagent teams. The proposed approach extends the cooperative coevolution architecture with operators that put the number of coevolving populations under evolutionary control. Populations are dynamically merged based on behavioral similarity, thus decreasing team heterogeneity, and stochastic population splits are used to explore increased team heterogeneity. Hyb-CCEA is capable of converging to suitable team compositions for the given task, be it a completely homogeneous team where all agents share the same control logic, a heterogeneous team where each agent has distinct control logic, or a partially heterogeneous team. By placing both team composition and agent controllers under evolutionary control, Hyb-CCEA can be applied to domains for which the experimenter has limited or no knowledge about possible solutions. We study Hyb-CCEA extensively in an abstract domain, and conduct a series of validation experiments with four simulated multirobot tasks: two multirover foraging tasks and two robotic soccer tasks. The results show that Hyb-CCEA takes advantage of partial heterogeneity and frequently outperforms the standard cooperative coevolution approach, both in terms of fitness scores achieved and number of evaluations needed to evolve solutions.

Cooperative coevolutionary algorithms (CCEAs) rely on multiple coevolving populations for the evolution of solutions composed of coadapted components. CCEAs enable, for instance, the evolution of cooperative multiagent systems composed of heterogeneous agents, where each agent is modelled as a component of the solution. Previous works have, however, shown that CCEAs are biased toward stability: the evolutionary process tends to converge prematurely to stable states instead of (near-)optimal solutions. In this study, we show how novelty search can be used to avoid the counterproductive attraction to stable states in coevolution. Novelty search is an evolutionary technique that drives evolution toward behavioural novelty and diversity rather than exclusively pursuing a static objective. We evaluate three novelty-based approaches that rely on, respectively (1) the novelty of the team as a whole, (2) the novelty of the agents' individual behaviour, and (3) the combination of the two. We compare the proposed approaches with traditional fitness-driven cooperative coevolution in three simulated multirobot tasks. Our results show that team-level novelty scoring is the most effective approach, significantly outperforming fitness-driven coevolution at multiple levels. Novelty-driven cooperative coevolution can substantially increase the potential of CCEAs while maintaining a computational complexity that scales well with the number of populations.

A long-standing goal in artificial intelligence and robotics is synthesising agents that can effectively learn and adapt throughout their lifetime. One open-ended approach to behaviour learning in autonomous robots is online evolution , which is part of the evolutionary robotics field of research. In online evolution approaches, an evolutionary algorithm is executed on the robots during task execution, which enables continuous optimisation and adaptation of behaviour. Despite the potential for automatic behaviour learning, online evolution has not been widely adopted because it often requires several hours or days to synthesise solutions to a given task. In this respect, research in the field has failed to develop a prevalent algorithm able to effectively synthesise solutions to a large number of different tasks in a timely manner. Rather than focusing on a single algorithm, we argue for more general mechanisms that can combine the benefits of different algorithms to increase the performance of online evolution of robot controllers. We conduct a comprehensive assessment of a novel approach called online hyper-evolution (OHE). Robots executing OHE use the different sources of feedback information traditionally associated with controller evaluation to find effective evolutionary algorithms during task execution. First, we study two approaches: OHE-fitness, which uses the fitness score of controllers as the criterion to select promising algorithms over time, and OHE-diversity, which relies on the behavioural diversity of controllers for algorithm selection. We then propose a novel class of techniques called OHE-hybrid, which combine diversity and fitness to search for suitable algorithms. In addition to their effectiveness at selecting suitable algorithms, the different OHE approaches are evaluated for their ability to construct algorithms by controlling which algorithmic components should be employed for controller generation (e.g., mutation, crossover, among others), an unprecedented approach in evolutionary robotics. Results show that OHE (i) facilitates the evolution of controllers with high performance, (ii) can increase effectiveness at different stages of evolution by combining the benefits of multiple algorithms over time, and (iii) can be effectively applied to construct new algorithms during task execution. Overall, our study shows that OHE is a powerful new paradigm that allows robots to improve their learning process as they operate in the task environment.

The use of evolutionary robotics in robots with complex means of locomotion has, so far, mainly been limited to gait evolution. Increasing the number of degrees of freedom available to a controller significantly enlarges the search space, which in turn makes the evolution of solutions for a given task more challenging. In this paper, we present Evolutionary Repertoire-based Control (EvoRBC), an approach that enables the evolution of control for robots with arbitrary locomotion complexity. EvoRBC separates the synthesis of control into two levels: the generation of a repertoire of behavior primitives through the application of Quality Diversity techniques; and the evolution of a behavior arbitrator that uses the repertoire's primitives to solve a particular task. We evaluate EvoRBC in simulated robots with different numbers of degrees of freedom in two tasks, navigation and foraging. Our results show that while standard evolutionary approaches are highly affected by the locomotion complexity of the robot, EvoRBC is consistently able to evolve high-performing solutions. We also show that EvoRBC allows for the evolution of general controllers, that can be successfully used in robots different than those with which they were evolved.

This paper presents OpenSwarm, a lightweight easy-to-use open-source operating system. To our knowledge, it is the first operating system designed for and deployed on miniature robots. OpenSwarm operates directly on a robot's microcontroller. It has a memory footprint of 1 kB RAM and 12 kB ROM. OpenSwarm enables a robot to execute multiple processes simultaneously. It provides a hybrid kernel that natively supports preemptive and cooperative scheduling, making it suitable for both computationally intensive and swiftly responsive robotics tasks. OpenSwarm provides hardware abstractions to rapidly develop and test platform-independent code. We show how OpenSwarm can be used to solve a canonical problem in swarm robotics—clustering a collection of dispersed objects. We report experiments, conducted with five e-puck mobile robots, that show that an OpenSwarm implementation performs as good as a hardware-near implementation. The primary goal of OpenSwarm is to make robots with severely constrained hardware more accessible, which may help such systems to be deployed in real-world applications.

One of the long-term goals in evolutionary robotics is to be able to automatically synthesize controllers for real autonomous robots based only on a task specification. While a number of studies have shown the applicability of evolutionary robotics techniques for the synthesis of behavioral control, researchers have consistently been faced with a number of issues preventing the widespread adoption of evolutionary robotics for engineering purposes. In this article, we review and discuss the open issues in evolutionary robotics. First, we analyze the benefits and challenges of simulation-based evolution and subsequent deployment of controllers versus evolution on real robotic hardware. Second, we discuss specific evolutionary computation issues that have plagued evolutionary robotics: (1) the bootstrap problem, (2) deception, and (3) the role of genomic encoding and genotype-phenotype mapping in the evolution of controllers for complex tasks. Finally, we address the absence of standard research practices in the field. We also discuss promising avenues of research. Our underlying motivation is the reduction of the current gap between evolutionary robotics and mainstream robotics, and the establishment of evolutionary robotics as a canonical approach for the engineering of autonomous robots.

Novelty search is a state-of-the-art evolutionary approach that promotes behavioural novelty instead of pursuing a static objective. Along with a large number of successful applications, many different variants of novelty search have been proposed. It is still unclear, however, how some key parameters and algorithmic components influence the evolutionary dynamics and performance of novelty search. In this paper, we conduct a comprehensive empirical study focused on novelty search's algorithmic components. We study the "k" parameter -- the number of nearest neighbours used in the computation of novelty scores; the use and function of an archive; how to combine novelty search with fitness-based evolution; and how to configure the mutation rate of the underlying evolutionary algorithm. Our study is conducted in a simulated maze navigation task. Our results show that the configuration of novelty search can have a significant impact on performance and behaviour space exploration. We conclude with a number of guidelines for the implementation and configuration of novelty search, which should help future practitioners to apply novelty search more effectively.

In this paper, we discuss wireless sensor and networking technologies for swarms of inexpensive aquatic surface drones in the context of the HANCAD project. The goal is to enable the swarm to perform maritime tasks such as sea-border patrolling and environmental monitoring, while keeping the cost of each drone low. Communication between drones is essential for the success of the project. Preliminary experiments show that XBee modules are promising for energy efficient multi-hop drone-to-drone communication.

Online evolution gives robots the capacity to learn new tasks and to adapt to changing environmental conditions during task execution. Previous approaches to online evolution of neural controllers are typically limited to the optimisation of weights in networks with a prespecified, fixed topology. In this article, we propose a novel approach to online learning in groups of autonomous robots called odNEAT. odNEAT is a distributed and decentralised neuroevolution algorithm that evolves both weights and network topology. We demonstrate odNEAT in three multirobot tasks: aggregation, integrated navigation and obstacle avoidance, and phototaxis. Results show that odNEAT approximates the performance of rtNEAT, an efficient centralised method, and outperforms IM-(μ + 1), a decentralised neuroevolution algorithm. Compared with rtNEAT and IM-(μ + 1), odNEAT's evolutionary dynamics lead to the synthesis of less complex neural controllers with superior generalisation capabilities. We show that robots executing odNEAT can display a high degree of fault tolerance as they are able to adapt and learn new behaviours in the presence of faults. We conclude with a series of ablation studies to analyse the impact of each algorithmic component on performance.

In this paper, we discuss how the combination of evolutionary techniques and engineering-oriented approaches is an effective methodology for leveraging the potential of evolutionary robotics (ER) in the synthesis of behavioural control. We argue that such combination can eliminate the issues that have prevented ER from scaling to complex real-world tasks,namely: (i) the bootstrap problem, (ii) deception, (iii) the reality gap effect, and (iv) the prohibitive amount of time necessary to evolve controllers directly in real robotic hardware. We present recent studies carried out in our research group involving real-robot and simulation-based experiments. We provide examples of how the synergistic effects of evolution and engineering overcome each other’s limitations and significantly extend their respective capabilities, thereby opening a new path in the design of robot controllers.

This paper introduces JBotEvolver, a versatile simulation platform for research and education in evolutionary robotics (ER). JBotEvolver is a Java-based open-source, cross-platform framework available at https://code. google.com/p/jbotevolver/ under the GNU GPL. JBotEvolver has been used in a number of previous ER studies of our research group, from offline evolution to online evolution and learning, and from single to multirobot systems (for examples see Duarte et al. (2014a,b); Silva et al. (2012a,b)), and in a number of undergraduate and graduate courses at ISCTE-IUL

Decentralized control for multirobot systems is difficult to design by hand because the behavioral rules for individual robots cannot, in general, be derived from a desired collective behavior. System designers have therefore resorted to evo-lutionary computation as a means to heuristically synthesize self-organized behaviors for robot collectives. Evolutionary computation is typically applied by putting the rules gov-erning the individual robots under evolutionary control and by assigning fitness scores based on collective performance. Scaling evolutionary approaches to complex tasks has, how-ever, proven challenging due to issues related to bootstrap-ping and premature convergence. In this paper, we show how hierarchical task decomposition and the combination of evolved and preprogrammed control can overcome these is-sues. We apply our approach to a complex multirobot task that requires a high degree of coordination and collective de-cision making, and we synthesize controllers capable of solv-ing the task.

Maritime tasks, such as surveillance and patrolling, aquaculture inspection, and wildlife monitoring, typically require large operational crews and expensive equipment. Only recently have unmanned vehicles started to be used for such missions. These vehicles, however, tend to be expensive and have limited coverage, which prevents large-scale deployment. In this paper, we propose a scalable robotics system based on swarms of small and inexpensive aquatic drones. We take advantage of bio-inspired artificial evolution techniques in order to synthesize scalable and robust collective behaviors for the drones. The behaviors are then combined hierarchically with preprogrammed control in an engineeredcentric approach, allowing the overall behavior for a particular mission to be quickly configured and tested in simulation before the aquatic drones are deployed. We demonstrate the scalability of our hybrid approach by successfully deploying up to 1,000 simulated drones to patrol a 20 km long strip for 24 hours.

Evolutionary techniques driven by behavioural diversity, such as novelty search, have shown significant potential in evolutionary robotics. These techniques rely on priorly specified behaviour characterisations to estimate the similarity between individuals. Characterisations are typically defined in an ad hoc manner based on the experimenter's intuition and knowledge about the task. Alternatively, generic characterisations based on the sensor-effector values of the agents are used. In this paper, we propose a novel approach that allows for systematic derivation of behaviour characterisations for evolutionary robotics, based on a formal description of the agents and their environment. Systematically derived behaviour characterisations (SDBCs) go beyond generic characterisations in that they can contain task-specific features related to the internal state of the agents, environmental features, and relations between them. We evaluate SDBCs with novelty search in three simulated collective robotics tasks. Our results show that SDBCs yield a performance comparable to the task-specific characterisations, in terms of both solution quality and behaviour space exploration.

The engineering of fault-detection systems for multirobot systems (MRS) is a well-studied problem (e.g, Christensen et al. (2009)). Most fault-detection models are built on the assumption that normal behavior is known, and can be characterized in advance. The models are trained to recognize predefined normal behaviors, and behaviors not recognized are labeled abnormal. While such an approach does provide some interesting results of robust fault detection and fault tolerance, they may not be applicable when normal behavior can change as a result of unforeseen environmental conditions and online learning for instance. Furthermore, prior information required to characterize normal behaviors, may not always be available.

Novelty search has shown to be a promising approach for the evolution of controllers for swarms of robots. In existing studies, however, the experimenter had to craft a task-specific behaviour similarity measure. The reliance on hand-crafted similarity measures places an additional burden to the experimenter and introduces a bias in the evolutionary process. In this paper, we propose and compare two generic behaviour similarity measures: combined state count and sampled average state. The proposed measures are based on the values of sensors and effectors recorded for each individual robot of the swarm. The characterisation of the group-level behaviour is then obtained by combining the sensor-effector values from all the robots. We evaluate the proposed measures in an aggregation task and in a resource sharing task. We show that the generic measures match the performance of task-specific measures in terms of solution quality. Our results indicate that the proposed generic measures operate as effective behaviour similarity measures, and that it is possible to leverage the benefits of novelty search without having to craft task-specific similarity measures.

We propose and evaluate a novel approach called Online Distributed NeuroEvolution of Augmenting Topologies (odNEAT). odNEAT is a completely distributed evolutionary algorithm for online learning in groups of embodied agents such as robots. While previous approaches to online distributed evolution of neural controllers have been limited to the optimisation of weights, odNEAT evolves both weights and network topology. We demonstrate odNEAT through a series of simulation-based experiments in which a group of e-puck-like robots must perform an aggregation task. Our results show that robots are capable of evolving effective aggregation strategies and that sustainable behaviours evolve quickly. We show that odNEAT approximates the performance of rtNEAT, a similar but centralised method. We also analyse the contribution of each algorithmic component on the performance through a series of ablation studies.

Top-down causation has been suggested to occur at all scales of biological organization as a mechanism for explaining the hierarchy of structure and causation in living systems. Here we propose that a transition from bottom-up to top-down causation -- mediated by a reversal in the flow of information from lower to higher levels of organization, to that from higher to lower levels of organization -- is a driving force for most major evolutionary transitions. We suggest that many major evolutionary transitions might therefore be marked by a transition in causal structure. We use logistic growth as a toy model for demonstrating how such a transition can drive the emergence of collective behavior in replicative systems. We then outline how this scenario may have played out in those major evolutionary transitions in which new, higher levels of organization emerged, and propose possible methods via which our hypothesis might be tested.

We introduce spatially targeted communication - a communication method for multirobot systems. This method allows an individual message sending robot to isolate selected message recipient robots based on their spatial location. The recipient robots can then be sent information targeted solely at them, even if the sending robot uses a broadcast communication modality. We demonstrate spatially targeted communication using a heterogeneous multirobot system composed of flying robots and ground-based self-assembling robots. Flying robots use their privileged view of the environment to determine and communicate information to groups of ground-based robots on what morphologies to form to carry out upcoming tasks.

A self-reconfigurable robot is a robot that is made up of a number of physically connected modules that can rearrange their connections to change the global shape of the robot. Given their morphological flexibility, self-reconfigurable robots have the potential to display a high degree of versatility: to move fast on flat terrain, modules may reconfigure into a circular shape and roll; to move in confined spaces, the modules may reconfigure into a snake and creep; and to manipulate an object, the modules may reconfigure into a shape specialized to the characteristics of the object and/or to the type of manipulation required.Research on self-reconfigurable robots started in the late 1980s when the idea of cellular robots emerged [3]. The vision was to develop a system of autonomous cells that could jointly form different shapes to accomplish tasks. Over the past two decades, significant progress has been made and several self-reconfigurable robots have been built.The book Self-Reconfigurable Robots—An Introduction by Stoy, Brandt, and Christensen is the first book on self-reconfigurable robots to appear. The authors have divided the material into 10 chapters covering topics ranging from the history of self-reconfigurable robotics and hardware to different approaches to self-reconfiguration and future research challenges. The book is aimed at graduate students and researchers. The material is, however, presented in language and at a level of technical detail that allow anyone with an interest in self-reconfigurable robots to enjoy the book.Chapter 1 introduces the reader to the field of self-reconfigurable robotics by providing two possible scenarios in which such robots could be particularly useful: (i) a planetary exploration scenario, in which a self-reconfigurable robot's morphological flexibility enables it to navigate in different environments, and (ii) a morphing production line scenario, in which reconfigurable robots autonomously assemble furniture. The authors then present examples of two self-reconfigurable robots, namely the CONRO [2] and the ATRON [5]. A brief history of self-reconfigurable robots is then provided. The history starts at Fukuda's CEBOT [3] concept and Yim's PolyPod [7]; some of the more recent systems covered include Shen et al.'s SuperBot [6] and Zykov et al.'s Molecubes [8]. The authors propose an interesting classification for self-reconfigurable robotic systems, based on the number of modules in the system: pack robots consisting of tens of modules, herd robots consisting of hundreds of modules, and swarm robots consisting of very large numbers of modules. The classification is justified by two observations: (i) systems of different sizes tend to be suitable for different tasks, and (ii) different types of control are appropriate for systems of different sizes.Chapter 2 is titled “Designing Self-Reconfigurable Robots.” In this chapter, the authors stress the interdependences between hardware, control, morphology, task(s), and environment(s). Some of the tradeoffs that have to be made in the design of self-reconfigurable robotic systems are also discussed. A distinction is made between design goals and characteristics of systems. Versatility, adaptability, robustness, and cheapness are listed as desirable design goals for most robotic systems, whereas systems can be characterized according to their degrees of reconfigurability, scalability, and responsiveness, and how well they meet the functional requirements in their domain of application. Although the distinction between design goals and characteristics seems slightly arbitrary (systems could arguably be categorized according to their degree of robustness, while scalability could be a design goal for some systems, for instance), the discussion of the individual items is relevant and provides valuable insight to system designers.Chapter 3 focuses on the mechanical design of self-reconfigurable robotic systems. The purpose of the chapter, as stated by the authors, is to “enlighten software designers about the cost of implementing a solution at the hardware level … [and] to provide the hardware designer with a starting point for designing self-reconfigurable robots.” The chapter starts off with a discussion of the different types of self-reconfigurable robots, namely chain-type, lattice-type, and hybrids. The authors introduce different aspects of mechanical design: reconfiguration in two dimensions and in three dimensions, module geometry, module autonomy, the use of submodules, heterogeneous modules, bipartite designs, and actuator strength. Different types of connectors—magnetic connectors, mechanical connectors, and electrostatic connectors—are also discussed.Chapter 3 spans 42 pages and is more than twice as long as most of the other nine chapters in the book. The space is devoted to an overview of mechanical design and its many facets rather than to low-level technical details. The chapter provides a good overview of existing designs and systems, and it discusses several of the considerations and tradeoffs that must be made in the mechanical design of self-reconfigurable robotic systems. A table of 21 self-reconfigurable robots, with the number and the type of connectors used in each system and references to the publication(s) on the respective systems, is provided. For readers who wish to dive into more technical details, the authors conveniently list 33 categorized references to publications covering the topics discussed in the chapter.Chapter 4 is titled “Electrical Design of Self-Reconfigurable Robots.” The majority of the chapter is devoted to a discussion of communication and to discussions of the advantages and disadvantages of different types of communication: local communication, global communication, and multimode communication. The authors briefly debate the use of external power supplies versus on-board batteries. A table summarizing the computational hardware, type of communication technology, and on-board sensors for 20 different self-reconfigurable robots is also provided.Chapter 5 is an introduction to the problem of self-reconfiguration. Self-reconfiguration is difficult for several reasons: (i) modules are often subject to motion constraints, (ii) modules can get trapped inside or outside hollow substructures, (iii) a robot usually has to remain in one piece during reconfiguration, and (iv) when multiple modules move at the same time, congestion can arise. An overview of approaches to simplify the problem of self-reconfiguration is then provided, namely (i) the use of metamodules, that is, larger modules that are composed of a number of smaller modules, (ii) reducing the size of the space of global configurations by only allowing for a limited set of local configurations, (iii) the use of scaffolds of modules on which other modules can move, and (iv) the use of a few predefined intermediate configurations during reconfiguration. For each of the simplification approaches, the authors discuss why the approach simplifies the reconfiguration problem and at what cost. The authors then consider the problem of self-reconfiguration from three different perspectives: self-reconfiguration as a search problem (Chapter 6), self-reconfiguration as a control problem (Chapter 7), and self-reconfiguration as a by-product of task execution (Chapter 8).In Chapter 6, the problem of how to find a sequence of module moves that will reconfigure the robot from an initial configuration to a goal configuration is discussed. Configurations of a robot can be represented as nodes in a graph while edges between nodes represent the move (or moves) necessary to go from one configuration to another. The self-reconfiguration problem as seen from a search perspective essentially boils down to finding a path from the node presenting the initial configuration to the node representing the goal configuration. However, as the authors demonstrate, the exponential growth in the size of the search space makes brute-force approaches infeasible—even for systems composed of relatively few modules. Various heuristics and distance metrics are discussed. The authors give examples of how simplifications, such as metamodules and scaffolds, have been used in order to make search a feasible approach to the problem of self-reconfiguration. As noted at the end of Chapter 6, the use of search often necessitates centralized control and a global view of the system.Self-reconfiguration viewed as a distributed control problem is the topic of Chapter 7. In the discussion, the authors assume that modules know their current location and their goal location with respect to the other modules. The challenge for each module is to determine how and where to move next in order to close the distance to its goal location. Different movement strategies are discussed: random movements, local rules, coordinate attractors, gradient attractors, and recruitment. The authors also discuss different representations of the goal locations and how a target shape can be grown using transition rules.Chapter 8 is on the topic of self-reconfiguration as a side effect of task execution. The authors discuss several simulation-based works on cluster flow. In cluster flow, a robot as a whole moves forward by continuously letting modules from the back of the robot wander to the front. The authors then present Bojinov et al.'s simulation-based work on task-driven self-reconfiguration through growth [1] and Ishiguro et al.'s work on reconfiguration as a side effect of module oscillation [4]. Toward the end of Chapter 8, the authors discuss some unaddressed challenges in self-reconfigurable robotics, namely, for robots to maintain balance during self-reconfiguration in environments with gravity, and reconfiguration in robots composed of heterogeneous modules.Chapter 9 discusses control of self-reconfigurable robots in fixed configurations. The focus is mainly on locomotion, but one page is devoted to a discussion of issues associated with manipulation. The authors discuss some of the approaches that have been proposed and studied—namely, gate control tables, hormone-based control, and role-based control—and the challenges related to distributed control. The PolyPod and the CONRO are used as examples.Chapter 10 is titled “Research Challenges,” and the authors discuss some of the main challenges that need to be overcome before self-reconfigurable robots can be used in real-world applications. It is argued that while we may soon see systems of up to tens of modules (pack robots) outside of labs, a number of fundamental research questions, such as how to program and control large systems, still have to be addressed before self-reconfigurable swarm systems can take on real-world tasks. The focus of the chapter is therefore on challenges for systems of tens of modules (pack robots) and for systems of hundreds of modules (herd robots). After a discussion of the complexities associated with carrying out real-world tasks, the authors advocate a transition toward a common framework for the control of self-reconfigurable robots. Under such a framework, it should be possible to discuss how to move from isolated low-level behaviors, such as a gait, to controllers capable of solving more advanced tasks. The authors outline what such a framework could be: a behavior-based-like framework that takes the distributed and modular nature self-reconfigurable robots into account. The authors discuss the role of basic behaviors, behavior adaptation, behavior selection, and a novel concept called the behavior mode. A behavior mode corresponds to a set of basic behaviors and a robot shape. A switch from one behavior mode to another causes the robot to execute a new set of basic behaviors and triggers self-reconfiguration. In this way, a complex task may be divided into a number of simpler subtasks, for each of which a dedicated behavior mode can be designed. The authors end the book by stating that this is an exciting time for self-reconfigurable robots because several fundamental problems have been solved and because self-reconfigurable robots may soon be used in real-world scenarios.Self-Reconfigurable Robots—An Introduction is a well-structured and easily comprehensible introduction to the field of self-reconfigurable robots. The book is the first of its kind, and it brings together much of the research related to self-reconfigurable robots that has taken place over the past two decades in a single, coherent, introductory text. The book covers a lot of ground—from module geometry to distributed control of complex gaits. Each chapter contains a “Further Reading” section with categorized references to publications on the topics and systems discussed in the chapter. Both beginners and experienced researchers will therefore find the book a valuable addition to their personal library.

We introduce enhanced directional self-assembly (EDSA) - a novel mechanism for morphology growth through the creation of directed connections in a self-assembling multirobot system. In our approach, a robot inviting a physical connection actively recruits the best located neighboring robot and guides the recruit to the location on its chassis where the connection is required. The proposed mechanism relies on local, high-speed communication between connection inviting robots and their recruits. Communication is based on a hybrid technology that combines radio and infrared to provide local relative positioning information when messages are transmitted between adjacent robots. Experiments with real robotic hardware show that EDSA is precise (misalignment of only 1.2° on average), robust (100% success rate for the experiments in this study) and fast (16.1 seconds on average from a distance of 80 cm). We show how the speed and precision of the new approach enable adaptive recruitment and connection in dynamic environments, a high degree of parallelism, and growth of a moving morphology.

This research work illustrates an approach to the design of controllers for self-assembling robots in which the self-assembly is initiated and regulated by perceptual cues that are brought forth by the physical robots through their dynamical interactions. More specifically, we present a homogeneous control system that can achieve assembly between two modules (two fully autonomous robots) of a mobile self-reconfigurable system without a priori introduced behavioral or morphological heterogeneities. The controllers are dynamic neural networks evolved in simulation that directly control all the actuators of the two robots. The neurocontrollers cause the dynamic specialization of the robots by allocating roles between them based solely on their interaction. We show that the best evolved controller proves to be successful when tested on a real hardware platform, the swarm-bot. The performance achieved is similar to the one achieved by existing modular or behavior-based approaches, also due to the effect of an emergent recovery mechanism that was neither explicitly rewarded by the fitness function, nor observed during the evolutionary simulation. Our results suggest that direct access to the orientations or intentions of the other agents is not a necessary condition for robot coordination: Our robots coordinate without direct or explicit communication, contrary to what is assumed by most research works in collective robotics. This work also contributes to strengthening the evidence that evolutionary robotics is a design methodology that can tackle real-world tasks demanding fine sensory-motor coordination.

One of the essential benefits of swarm robotic systems is redundancy. In case one robot breaks down, another robot can take steps to repair the failed robot or take over the failed robot's task. Although fault tolerance and robustness to individual failures have often been central arguments in favor of swarm robotic systems, few studies have been dedicated to the subject. In this paper, we take inspiration from the synchronized flashing behavior observed in some species of fireflies. We derive a completely decentralized algorithm to detect non-operational robots in a swarm robotic system. Each robot flashes by lighting up its on-board light-emitting diodes (LEDs), and neighboring robots are driven to flash in synchrony. Since robots that are suffering catastrophic failures do not flash periodically, they can be detected by operational robots. We explore the performance of the proposed algorithm both on a real-world swarm robotic system and in simulation. We show that failed robots are detected correctly and in a timely manner, and we show that a system composed of robots with simulated self-repair capabilities can survive relatively high failure rates.

In this paper, we propose SWARMORPH: a distributed morphology generation mechanism for autonomous self-assembling mobile robots. Self-organized growth of global morphological structures emerges through the repeated application of local morphology extension rules. We present details of the directional self-assembly mechanism that provides control over the orientation of interrobot connections. We conduct real-world experiments to validate the low-level directional self-assembly mechanism and the growth of global morphologies. We demonstrate the scalability of the approach with large numbers of robots in simulation-based experiments.

In this study, we show a group of robots can synchronize based on firefly-inspired flashing behavior and how dead robots can be detected by other robots. The algorithm is completely distributed. Each robot flashes by lighting up its on-board LEDs and neighboring robots are driven to flash in synchrony. Since robots that are suffering catastrophic failures do not flash periodically, they can be detected by operational robots. On a real multi-robot system of 10 autonomous robots, we show how the group can correctly detect multiple faults, and that when given (simulated) repair capabilities, the group can survive a relatively high rate of failure.

This paper proposes a distributed control mechanism for a self-propelled self-assembling robotic system that allows robots to form specific, connected morphologies. Global morphologies are 'grown' using local visual perception only. The robots in the system do not have access to a blueprint of the global pattern and the algorithmic rules are solely based on what a single robot can see in its immediate surroundings. None of the robots have any predefined position in the final morphology, except for the seed robot that initiates the self-assembly process. Robots that are part of the connected entity indicate where new robots should attach in order to grow the local structure appropriately. The paper demonstrates the efficacy of the mechanism by letting groups of up to 9 real robots self-assemble into four different morphologies: line, star, arrow, and dense.

Mobile robots are said to be capable of self- assembly when they can autonomously form physical connections with each other. Despite the recent proliferation of self- assembling systems, little work has been done on using self- assembly to add functional value to a robotic system, and even less on quantifying the contribution of self-assembly to system performance. In this study we demonstrate and quantify the performance benefits of i) acting as a physically larger self-assembled entity, ii) using self-assembly adaptively and iii) making the robots morphologically aware (the self-assembled robots leverage their new connected morphology in a task specific way). In our experiments, two real robots must navigate to a target over a-priori unknown terrain. In some cases the terrain can only be overcome by a self-assembled connected entity. In other cases, the robots can reach the target faster by navigating individually.