Search

There is a growing demand for the use of robots to assist humans in their tasks, especially those involving risks, such as search and rescue. For this reason, coordination among several robots has been a common option, and one of the ways to study and model these applications involves the problem of pursuit evasion. This paper extends the results presented earlier on the use of an evolutionary robotics approach to solve the worst case pursuit-evasion problem, in which evaders are considered arbitrarily fast and omniscient, while pursuers have limited sensing and communication capabilities, with no prior knowledge regarding environments, which are treated as discrete and can be multiply connected. First, a formulation based on random walk is offered. Then, the concept is extended to include a decentralized multi-robot control system based on a finite-state machine with state-action mapping defined by means of a genetic algorithm. Results show that the proposed system is able to decontaminate several types of maps, but does not generalize to all initial conditions, due to the incompleteness in the automaton mapping. Therefore, a complementary approach is presented in which random walk is used alternatively with the evolved automaton, indicating random actions in cases of states not sufficiently visited during evolution. In addition, a comparative analysis of the evolutionary approach and the random walk formulation is also carried out.

Complexity measures and information theory metrics in general have recently been attracting the interest of multi-agent and robotics communities, owing to their capability of capturing relevant features of robot behaviors, while abstracting from implementation details. We believe that theories and tools from complex systems science and information theory may be fruitfully applied in the near future to support the automatic design of robot swarms and the analysis of their dynamics. In this paper we discuss opportunities and open questions in this scenario.

Designing collective behaviors for robot swarms is a difficult endeavor due to their fully distributed, highly redundant, and ever-changing nature. To overcome the challenge, a few approaches have been proposed, which can be classified as manual, semi-automatic, or automatic design. This paper is intended to be the manifesto of the automatic off-line design for robot swarms. We define the off-line design problem and illustrate it via a possible practical realization, highlight the core research questions, raise a number of issues regarding the existing literature that is relevant to the automatic off-line design, and provide guidelines that we deem necessary for a healthy development of the domain and for ensuring its relevance to potential real-world applications.

Robot-to-robot learning, a specific case of social learning in robotics, enables multiple robots to share learned skills while completing a task. The literature offers various statements of its benefits. Robots using this type of social learning can reach a higher performance, an increased learning speed, or both, compared to robots using individual learning only. No general explanation has been advanced for the difference in observations, which make the results highly dependent on the particular system and parameter setting. In this paper, we perform a detailed analysis into the effects of robot-to-robot learning. As a result, we show that this type of social learning can reduce the sensitivity of the learning process to the choice of parameters in two ways. First, robot-to-robot learning can reduce the number of bad performing individuals in the population. Second, robot-to-robot learning can increase the chance of having a successful run, where success is defined as the presence of a high performing individual. Additionally, we show that robot-to-robot learning results in an increased learning speed for almost all parameter settings. Our results indicate that robot-to-robot learning is a powerful mechanism which leads to benefits in both performance and learning speed.

The concept of social interaction is at the core of embodied and enactive approaches to social cognitive processes, yet scientifically it remains poorly understood. Traditionally, cognitive science had relegated all behavior to being the end result of internal neural activity. However, the role of feedback from the interactions between agent and their environment has become increasingly important to understanding behavior. We focus on the role that social interaction plays in the behavioral and neural activity of the individuals taking part in it. Is social interaction merely a source of complex inputs to the individual, or can social interaction increase the individuals' own complexity? Here we provide a proof of concept of the latter possibility by artificially evolving pairs of simulated mobile robots to increase their neural complexity, which consistently gave rise to strategies that take advantage of their capacity for interaction. We found that during social interaction, the neural controllers exhibited dynamics of higher-dimensionality than were possible in social isolation. Moreover, by testing evolved strategies against unresponsive ghost partners, we demonstrated that under some conditions this effect was dependent on mutually responsive co-regulation, rather than on the mere presence of another agent's behavior as such. Our findings provide an illustration of how social interaction can augment the internal degrees of freedom of individuals who are actively engaged in participation.

We study evolutionary robot systems where not only the robot brains but also the robot bodies are evolvable. Such systems need to include a learning period right after ‘birth' to acquire a controller that fits the newly created body. In this paper we investigate the possibility of bootstrapping infant robot learning through employing Lamarckian inheritance of parental controllers. In our system controllers are encoded by a combination of a morphology dependent component, a Central Pattern Generator (CPG), and a morphology independent part, a Compositional Pattern Producing Network (CPPN). This makes it possible to transfer the CPPN part of controllers between different morphologies and to create a Lamarckian system. We conduct experiments with simulated modular robots whose fitness is determined by the speed of locomotion, establish the benefits of inheriting optimized parental controllers, shed light on the conditions that influence these benefits, and observe that changing the way controllers are evolved also impacts the evolved morphologies.

The Evolutionary Robotics (ER) process has been applied extensively to developing control programs to achieve locomotion in legged robots, as an automated alternative to the arduous task of manually creating control programs for such robots. The evolution of such controllers is typically performed in simulation by making use of a physics engine-based robotic simulator. Making use of such physics-based simulators does, however, have certain challenges associated with it, such as these simulators' computational inefficiency, potential issues with lack of accuracy and the human effort required to construct such simulators. The current study therefore proposed and investigated an alternative method of simulation for a hexapod (six-legged) robot in the ER process, and directly compared this newly-proposed simulation method to traditional physics-based simulation. This alternative robotic simulator was built based solely on experimental data acquired directly from observing the behaviour of the robot. This data was used to construct a simulator for the robot based on Artificial Neural Networks (ANNs). To compare this novel simulation method to traditional physics simulation, the ANN-based simulators were used to evolve simple open-loop locomotion controllers for the robot in simulation. The real-world performance of these controllers was compared to that of controllers evolved in a more traditional physics-based simulator. The obtained results indicated that the use of ANN-based simulators produced controllers which could successfully perform the required locomotion task on the real-world robot. In addition, the controllers evolved using the ANN-based simulators allowed the real-world robot to move further than those evolved in the physics-based simulator and the ANN-based simulators were vastly more computationally efficient than the physics-based simulator. This study thus decisively indicated that ANN-based simulators offer a superior alternative to widely-used physics simulators in ER for the locomotion task considered. [ABSTRACT FROM AUTHOR]

Engineers, control theorists, and neuroscientists often view the delay imposed by finite signal propagation velocities as a problem that needs to be compensated for or avoided. In this article, we consider the alternative possibility that in some cases, signal delay can be used functionally, that is, as an essential component of a cognitive system. To investigate this idea, we evolve a minimal robot controller to solve a basic stimulus-distinction task. The controller is constrained so that the solution must utilize a delayed recurrent signal. Different from previous evolutionary robotics studies, our controller is modeled using delay differential equations, which (unlike the ordinary differential equations of conventional continuous-time recurrent neural networks) can accurately capture delays in signal propagation. We analyze the evolved controller and its interaction with its environment using classical dynamical systems techniques. The analysis shows what kinds of invariant sets underlie the various successful and unsuccessful performances of the robot, and what kinds of bifurcations produce these invariant sets. In the second phase of our analysis, we turn our attention to the parameter θ, which describes the amount of signal delay included in the model. We show how the delay destabilizes certain attractors that would exist if there were no delay and creates other stable attractors, resulting in an agent that performs well at the target task. [ABSTRACT FROM AUTHOR]

This paper surveys various applications of artificial evolution in the field of modular robots. Evolutionary robotics aims to design autonomous adaptive robots automatically that can evolve to accomplish a specific task while adapting to environmental changes. A number of studies have demonstrated the feasibility of evolutionary algorithms for generating robotic control and morphology. However, a huge challenge faced was how to manufacture these robots. Therefore, modular robots were employed to simplify robotic evolution and their implementation in real hardware. Consequently, more research work has emerged on using evolutionary computation to design modular robots rather than using traditional hand design approaches in order to avoid cognition bias. These techniques have the potential of developing adaptive robots that can achieve tasks not fully understood by human designers. Furthermore, evolutionary algorithms were studied to generate global modular robotic behaviors including; self-assembly, self-reconfiguration, self-repair, and self-reproduction. These characteristics allow modular robots to explore unstructured and hazardous environments. In order to accomplish the aforementioned evolutionary modular robotic promises, this paper reviews current research on evolutionary robotics and modular robots. The motivation behind this work is to identify the most promising methods that can lead to developing autonomous adaptive robotic systems that require the minimum task related knowledge on the designer side. [ABSTRACT FROM AUTHOR]

A testbed has recently been introduced that evolves controllers for arbitrary hover-capable UAVs, with evaluations occurring directly on the robot. To prepare the testbed for real-world deployment, we investigate the effects of state-space limitations brought about by physical tethering (which prevents damage to the UAV during stochastic tuning), on the generality of the evolved controllers. We identify generalisation issues in some controllers, and propose an improved method that comprises two stages: in the first stage, controllers are evolved as normal using standard tethers, but experiments are terminated when the population displays basic flight competency. Optimisation then continues on a much less restrictive tether, effectively free-flying, and is allowed to explore a larger state-space envelope. We compare the two methods on a hover task using a real UAV, and show that more general solutions are generated in fewer generations using the two-stage approach. A secondary experiment undertakes a sensitivity analysis of the evolved controllers. [ABSTRACT FROM AUTHOR]

Many experiments have been performed that use evolutionary algorithms for learning the topology and connection weights of a neural network that controls a robot or virtual agent. These experiments are not only performed to better understand basic biological principles, but also with the hope that with further progress of the methods, they will become competitive for automatically creating robot behaviors of interest. However, current methods are limited with respect to the (Kolmogorov) complexity of evolved behavior. Using the evolution of robot trajectories as an example, we show that by adding four features, namely (1) freezing of previously evolved structure, (2) temporal scaffolding, (3) a homogeneous transfer function for output nodes, and (4) mutations that create new pathways to outputs, to standard methods for the evolution of neural networks, we can achieve an approximately linear growth of the complexity of behavior over thousands of generations. Overall, evolved complexity is up to two orders of magnitude over that achieved by standard methods in the experiments reported here, with the major limiting factor for further growth being the available run time. Thus, the set of methods proposed here promises to be a useful addition to various current neuroevolution methods. [ABSTRACT FROM AUTHOR]

This paper focuses on the effect of congestion on swarm performance by considering the number of robots and their size. Swarm robotics is the study of a large group of autonomous robots from which collective behavior emerges without reliance on any centralized control. Due to the fact that robotic swarms are composed of a large number of robots, it is important to consider the congestion among them. However, only a few studies have focused on the relationship between the congestion and the performance of robotic swarms;moreover, these studies only discuss the effect of the number of robots. In this study, experiments were conducted by computer simulation and carried out by varying both the number of robots and the size of the robots in a path formation task. The robot controller was designed with an evolutionary robotics approach. The results show that not only the number of robots but also their size are essential features in the relationship between congestion and swarm performance. In addition, autonomous specialization within the robotic swarm emerged in situations with moderate congestion. [ABSTRACT FROM AUTHOR]

Bipedal hopping is an efficient form of locomotion, yet it remains relatively rare in the natural world. Previous research has suggested that the tail balances the angular momentum of the legs to produce steady state bipedal hopping. In this study, we employ a 3D physics simulation engine to optimize gaits for an animat whose control and morphological characteristics are subject to computational evolution, which emulates properties of natural evolution. Results indicate that the order of gene fixation during the evolutionary process influences whether a bipedal hopping or quadrupedal bounding gait emerges. Furthermore, we found that in the most effective bipedal hoppers the tail balances the angular momentum of the torso, rather than the legs as previously thought. Finally, there appears to be a specific range of tail masses, as a proportion of total body mass, wherein the most effective bipedal hoppers evolve. [ABSTRACT FROM AUTHOR]

It has often been found that students appreciate hands-on work, and find that they learn more with courses that include a project than those relying solely on conventional lectures and tests. This type of project driven learning is a key component of "Inquiry-based learning" (IBL), which aims at teaching methodology as well as content by incorporating the student as an actor rather than a spectator. Robotics applications are especially well-suited for IBL due to the value of trial and error experience, the multiple possibilities for students to implement their own ideas, and the importance of programming, problem-solving, and electro-mechanical skills in real world engineering and science jobs. Furthermore, robotics platforms can be useful teaching media and learning tools for a variety of topics. Here, we present RoboGen: an open-source, web-based, software, and hardware platform for Robotics and Artificial Intelligence with a particular focus on Evolutionary Robotics. We describe the platform in detail, compare it to existing alternatives, and present results of its use as a platform for Inquiry-based learning within a master's level course at the Ecole Polytechnique Fédérale de Lausanne. [ABSTRACT FROM AUTHOR]

Abstract Novelty search is a tool in evolutionary and swarm robotics for maintaining the diversity of population needed for continuous robotic operation. It enables nature-inspired algorithms to evaluate solutions on the basis of the distance to their k -nearest neighbors in the search space. Besides this, the fitness function represents an additional measure for evaluating the solution, with the purpose of preserving the so-named novelty solutions into the next generation. In this study, a differential evolution was hybridized with novelty search. The differential evolution is a well-known algorithm for global optimization, which is applied to improve the results obtained by the other solvers on the CEC-14 benchmark function suite. Furthermore, functions of different dimensions were taken into consideration, and the influence of the various novelty search parameters was analyzed. The results of experiments show a great potential for using novelty search in global optimization. [ABSTRACT FROM AUTHOR]

The concept of social interaction is at the core of embodied and enactive approaches to social cognitive processes, yet scientifically it remains poorly understood. Traditionally, cognitive science had relegated all behavior to being the end result of internal neural activity. However, the role of feedback from the interactions between agent and their environment has become increasingly important to understanding behavior. We focus on the role that social interaction plays in the behavioral and neural activity of the individuals taking part in it. Is social interaction merely a source of complex inputs to the individual, or can social interaction increase the individuals' own complexity? Here we provide a proof of concept of the latter possibility by artificially evolving pairs of simulated mobile robots to increase their neural complexity, which consistently gave rise to strategies that take advantage of their capacity for interaction. We found that during social interaction, the neural controllers exhibited dynamics of higher-dimensionality than were possible in social isolation. Moreover, by testing evolved strategies against unresponsive ghost partners, we demonstrated that under some conditions this effect was dependent on mutually responsive co-regulation, rather than on the mere presence of another agent's behavior as such. Our findings provide an illustration of how social interaction can augment the internal degrees of freedom of individuals who are actively engaged in participation. [ABSTRACT FROM AUTHOR]

Heterogeneous multirobot systems have shown significant potential in many applications. Cooperative coevolutionary algorithms (CCEAs) represent a promising approach to synthesise controllers for such systems, as they can evolve multiple co-adapted components. Although CCEAs allow for an arbitrary level of team heterogeneity, in previous works heterogeneity is typically only addressed at the behavioural level. In this paper, we study the use of CCEAs to evolve control for a heterogeneous multirobot system where the robots have disparate morphologies and capabilities. Our experiments rely on a simulated task where a simple ground robot must cooperate with a complex aerial robot to find and collect items. We first show that CCEAs can evolve successful controllers for physically heterogeneous teams, but find that differences in the complexity of the skills the robots need to learn can impair CCEAs' effectiveness. We then study how different populations can use different evolutionary algorithms and parameters tuned to the agents' complexity. Finally, we demonstrate how CCEAs' effectiveness can be improved using incremental evolution or novelty-driven coevolution. Our study shows that, despite its limitations, coevolution is a viable approach for synthesising control for morphologically heterogeneous systems. [ABSTRACT FROM AUTHOR]

Swarm robotics is a field in which multiple robots coordinate their collective behavior autonomously to accomplish a given task without any form of centralized control. In swarm robotics, task allocation refers to the behavior resulting in robots being dynamically distributed over different sub-tasks, which is often required for solving complex tasks. It has been well recognized that evolutionary robotics is a promising approach to the development of collective behaviors for robotic swarms. However, the artificial evolution often suffers from two issues—the bootstrapping problem and deception—especially when the underlying task is profoundly complex. In this study, we propose a two-step scheme consisting of task partitioning and autonomous task allocation to overcome these difficulties. We conduct computer simulation experiments where robotic swarms have to accomplish a complex collective foraging problem, and the results show that the proposed approach leads to perform more effectively than a conventional evolutionary robotics approach. [ABSTRACT FROM AUTHOR]

Neurorobots are robots whose control has been modeled after some aspect of the brain. Since the brain is so closely coupled to the body and situated in the environment, Neurorobots can be a powerful tool for studying neural function in a holistic fashion. It may also be a means to develop autonomous systems that have some level of biological intelligence. The present article provides my perspective on this field, points out some of the landmark events, and discusses its future potential. [ABSTRACT FROM AUTHOR]

Irrigated agriculture is one of the key factors responsible for decreasing freshwater availability in recent years. Thus, the development of new tools which will help Irrigation District managers in their daily decision making process about the use of water and energy is essential. On the other hand, the new era of Big Data and information and communications technologies (ICT) has made it possible to have a larger amount of information available, leading to the development of new prediction tools. However, the quality and quantity of this information in many fields such as irrigated agriculture is limited. Consequently, the way in which the development of new predictive models is addressed must be reformulated. Thus, in this work, a new methodology to provide short-term forecasting of daily irrigation water demand when data availability is limited has been developed by coupling dynamic Artificial Neural Networks (ANN) architecture, the Bayesian framework and Genetic Algorithms (GA). The methodology was applied in the Bembézar MD Irrigation District (Southern Spain). The developed model improved the prediction accuracy by between 3% and 11% with respect to previous work. The best ANN model had a Standard Error Prediction (SEP) and a determination coefficient (R2) of 8.7% and 96%, respectively. The accuracy of the model developed makes it a powerful tool for the daily management of irrigation districts. Highlights • Forecasting of daily irrigation water demand model with data available is limited. • The Bayesian framework approach and GAs were used to find the optimal ANN. • The best model had a SEP and R2 values of 8.66% and 96%, respectively. • The methodology improves considerably the models developed by other authors. • The model developed provides new useful tools to the Irrigation District managers. [ABSTRACT FROM AUTHOR]

In order to be fully autonomous, robots have to be resilient so that they can recover from damages and operate for a long period of time with no human assistance. To be resilient, existing approaches propose to change the robots' behavior using a different control system when a hardware fault or damage occurs. These approaches are used for robots which have fixed morphologies. However, we cannot assume which morphology would be optimal for a given problem and which morphology allows resilience. In the present paper, we introduce a new approach that generates resilient artificial modular robots by evolving the robot morphology along with its controller. We used a multi-objective evolutionary algorithm to optimize two objectives at a time, which are the traveled distance of a damage-free robot and the traveled distance of the same robot with damaged parts. The result of preliminary experiments demonstrates that during evaluation, when robots are deliberately faced to motor failures, the evolution process can optimize and generate new morphologies for which the robot's behavior is less affected by damage. This makes the robot capable to recover its ability to move forward. [ABSTRACT FROM AUTHOR]

There are several ubiquitous kinematic structures that are used in industrial robots, with the most prominent being a six-axis angular structure. However, researchers are experimenting with task-based mechanism synthesis that could provide higher efficiency with custom optimized manipulators. Many studies have focused on finding the most efficient optimization algorithm for task-based robot manipulators. These manipulators, however, are usually optimized from simple modular joints and links, without exploring more elaborate modules. Here, we show that link modules defined by small numbers of parameters have better performance than more complicated ones. We compare four different manipulator link types, namely basic predefined links with fixed dimensions, straight links that can be optimized for different lengths, rounded links, and links with a curvature defined by a Hermite spline. Manipulators are then built from these modules using a genetic algorithm and are optimized for three different tasks. The results demonstrate that manipulators built from simple links not only converge faster, which is expected given the fewer optimized parameters, but also converge on lower cost values.

We investigate whether standard evolutionary robotics methods can be extended to support the evolution of multiple behaviors by forcing the retention of variations that are adaptive with respect to all required behaviors. This is realized by selecting the individuals located in the first Pareto fronts of the multidimensional fitness space in the case of a standard evolutionary algorithms and by computing and using multiple gradients of the expected fitness in the case of a modern evolutionary strategies that move the population in the direction of the gradient of the fitness. The results collected on two extended versions of state-of-the-art benchmarking problems indicate that the latter method permits to evolve robots capable of producing the required multiple behaviors in the majority of the replications and produces significantly better results than all the other methods considered.

A long-term vision of evolutionary robotics is a technology enabling the evolution of entire autonomous robotic ecosystems that live and work for long periods in challenging and dynamic environments without the need for direct human oversight. Evolutionary robotics has been widely used due to its capability of creating unique robot designs in simulation. Recent work has shown that it is possible to autonomously construct evolved designs in the physical domain; however, this brings new challenges: the autonomous manufacture and assembly process introduces new constraints that are not apparent in simulation. To tackle this, we introduce a new method for producing a repertoire of diverse but manufacturable robots. This repertoire is used to seed an evolutionary loop that subsequently evolves robot designs and controllers capable of solving a maze-navigation task. We show that compared to random initialisation, seeding with a diverse and manufacturable population speeds up convergence and on some tasks, increases performance, while maintaining manufacturability.

RealTimeBattle is an environment in which robots controlled by programs fight each other. Programs control the simulated robots using low-level messages (e.g., turn radar, accelerate). Unlike other tools like Robocode, each of these robots can be developed using different programming languages. Our purpose is to generate, without human programming or other intervention, a robot that is highly competitive in RealTimeBattle. To that end, we implemented an Evolutionary Computation technique: Genetic Programming. The robot controllers created in the course of the experiments exhibit several different and effective combat strategies such as avoidance, sniping, encircling and shooting. To further improve their performance, we propose a function-set that includes short-term memory mechanisms, which allowed us to evolve a robot that is superior to all of the rivals used for its training. The robot was also tested in a bout with the winner of the previous “RealTimeBattle Championship”, which it won. Finally, our robot was tested in a multi-robot battle arena, with five simultaneous opponents, and obtained the best results among the contenders.

This paper reflects on a seminal work in the history of AI and representation: Rodney Brooks’ 1991 paper Intelligence without representation. Brooks advocated the removal of explicit representations and engineered environments from the domain of his robotic intelligence experimentation, in favour of an evolutionary-inspired approach using layers of reactive behaviour that operated independently of each other. Brooks criticised the current progress in AI research and believed that removing complex representation from AI would help address problematic areas in modelling the mind. His belief was that we should develop artificial intelligence by being guided by the evolutionary development of our own intelligence and that his approach mirrored how our own intelligence functions. Thus, the field of behaviour-based robotics emerged. This paper offers a historical analysis of Brooks’ behaviour-based robotics approach and its impact on artificial intelligence and cognitive theory at the time, as well as on modern-day approaches to AI.

Research in swarm robotics has shown that automatic design is an effective approach to realize robot swarms. In automatic design methods, the collective behavior of a swarm is obtained by automatically configuring and fine-tuning the control software of individual robots. In this paper, we present TuttiFrutti: an automatic design method for robot swarms that belongs to AutoMoDe—a family of methods that produce control software by assembling preexisting software modules via optimization. The peculiarity of TuttiFrutti is that it designs control software for e-puck robots that can display and perceive colors using their RGB LEDs and omnidirectional camera. Studies with AutoMoDe have been so far restricted by the limited capabilities of the e-pucks. By enabling the use of colors, we significantly enlarge the variety of collective behaviors they can produce. We assess TuttiFrutti with swarms of e-pucks that perform missions in which they should react to colored light. Results show that TuttiFrutti designs collective behaviors in which the robots identify the colored light displayed in the environment and act accordingly. The control software designed by TuttiFrutti endowed the swarms of e-pucks with the ability to use color-based information for handling events, communicating, and navigating.

Neurorobots are robots whose control has been modeled after some aspect of the brain. Since the brain is so closely coupled to the body and situated in the environment, Neurorobots can be a powerful tool for studying neural function in a holistic fashion. It may also be a means to develop autonomous systems that have some level of biological intelligence. The present article provides my perspective on this field, points out some of the landmark events, and discusses its future potential.

We introduce the Xpuck swarm, a research platform with an aggregate raw processing power in excess of two teraflops. The swarm uses 16 e-puck robots augmented with custom hardware that uses the substantial CPU and GPU processing power available from modern mobile system-on-chip devices. The augmented robots, called Xpucks, have at least an order of magnitude greater performance than previous swarm robotics platforms. The platform enables new experiments that require high individual robot computation and multiple robots. Uses include online evolution or learning of swarm controllers, simulation for answering what-if questions about possible actions, distributed super-computing for mobile platforms, and real-world applications of swarm robotics that requires image processing, or SLAM. The teraflop swarm could also be used to explore swarming in nature by providing platforms with similar computational power as simple insects. We demonstrate the computational capability of the swarm by implementing a fast physics-based robot simulator and using this within a distributed island model evolutionary system, all hosted on the Xpucks.

This article provides an overview of evolutionary robotics techniques applied to online distributed evolution for robot collectives, namely, embodied evolution. It provides a definition of embodied evolution as well as a thorough description of the underlying concepts and mechanisms. This article also presents a comprehensive summary of research published in the field since its inception around the year 2000, providing various perspectives to identify the major trends. In particular, we identify a shift from considering embodied evolution as a parallel search method within small robot collectives (fewer than 10 robots) to embodied evolution as an online distributed learning method for designing collective behaviors in swarm-like collectives. This article concludes with a discussion of applications and open questions, providing a milestone for past and an inspiration for future research.