Truss Links are a modular robot platform originally developed for studying Robot Metabolism, the ability for machines to grow, repair, and intelligently adapt by integrating components from their environment or other robots. Here, we investigate whether Truss Links can be more than robots that build themselves: can they serve as a creative medium accessible to non-experts? In this paper, we present a preliminary research-through-design exploration of Truss Links as digital building blocks for the physical world. We demonstrate how Truss Links can be used to animate text, animate two- and three-dimensional shapes, form kinetic sculptures, and hand-build a delta pick-and-place robot. We discuss what this exploration reveals about the design properties that can make a research-grade robotic platform accessible as a creative medium.

Robot description models are essential for simulation and control, yet their creation often requires significant manual effort. To streamline this modeling process, we introduce AutoURDF, an unsupervised approach for constructing description files of unseen robots from point cloud frames. Our method leverages a cluster-based point cloud registration model that tracks the 6-DoF transformations of point clusters. Through analyzing cluster movements, we hierarchically address the following challenges: (1) moving part segmentation, (2) body topology inference, and (3) joint parameter estimation. The complete pipeline produces robot description files that are fully compatible with existing simulators. We validate our method across a variety of robots, using both synthetic and real-world scan data. Results indicate that our approach outperforms previous methods in registration and body topology estimation accuracy, offering a scalable solution for automated robot modeling.

A robot self-model is a task-agnostic representation of the robot’s physical morphology that can be used for motion planning tasks in the absence of a classical geometric kinematic model. In particular, when the latter is hard to engineer or the robot’s kinematics change unexpectedly, human-free self-modeling is a necessary feature of truly autonomous agents. In this work, we leverage neural fields to allow a robot to self-model its kinematics as a neural-implicit query model learned only from 2D images annotated with camera poses and configurations. This enables significantly greater applicability than existing approaches which have been dependent on depth images or geometry knowledge. To this end, alongside a curricular data sampling strategy, we propose a new encoder-based neural density field architecture for dynamic object-centric scenes conditioned on high numbers of degrees of freedom (DOFs). In a 7-DOF robot test setup, the learned self-model achieves a Chamfer-L2 distance of 2% of the robot’s workspace dimension. We demonstrate the capabilities of this model on motion planning tasks as an exemplary downstream application.

Integrating Large Language Models (LLMs) and Vision-Language Models (VLMs) with robotic systems enables robots to process and understand complex natural language instructions and visual information. However, a fundamental challenge remains: for robots to fully capitalize on these advancements, they must have a deep understanding of their physical embodiment. The gap between AI models’ cognitive capabilities and the understanding of physical embodiment leads to the following question: Can a robot autonomously understand and adapt to its physical form and functionalities through interaction with its environment? This question underscores the transition towards developing self-modeling robots without reliance on external sensory or pre-programmed knowledge about their structure. Here, we propose a meta-self-modeling that can deduce robot morphology through proprioception—the robot’s internal sense of its body’s position and movement. Our study introduces a 12-DoF reconfigurable legged robot, accompanied by a diverse dataset of 200k unique configurations, to systematically investigate the relationship between robotic motion and robot morphology. Utilizing a deep neural network model comprising a robot signature encoder and a configuration decoder, we demonstrate the capability of our system to accurately predict robot configurations from proprioceptive signals. This research contributes to the field of robotic self-modeling, aiming to enhance robot’s understanding of their physical embodiment and adaptability in real-world scenarios.

Enlightened by the fast-running gait of mammals like cheetahs and wolves, we design and fabricate a single-actuated untethered compliant robot that is capable of galloping at a speed of 313 mm/s or 1.56 body length per second (BL/s), faster than most reported soft crawlers in mm/s and BL/s. An in-plane prestressed hair clip mechanism (HCM) made up of semirigid materials, i.e. plastics are used as the supporting chassis, the compliant spine, and the force amplifier of the robot at the same time, enabling the robot to be simple, rapid, and strong. With experiments, we find that the HCM robotic locomotion speed is linearly related to actuation frequencies and substrate friction differences except for concrete surface, that tethering slows down the crawler, and that asymmetric actuation creates a new galloping gait. This paper demonstrates the potential of HCM-based soft robots.

Robot grasping is subject to an inherent tradeoff: Grippers with a large span typically take a longer time to close, and fast grippers usually cover a small span. However, many practical applications of grippers require the ability to close a large distance rapidly. For example, grasping cloth typically requires pressing a wide span of fabric into a graspable cusp. Besides, the ability to perform human-like grasping and ease offabrication are also very important for new soft grippers. Here, we demonstrate a human-finger-inspired snapping gripper that exploits elastic instability to achieve rapid and reversible closing over a wide span. Using prestressed semi-rigid material as the skeleton, the gripper fingers can widely open (86 mm) and rapidly close (46 ms) following a trajectory similar to that of a thumb-index finger pinching, and is 2.7 times and 10.9 times better than the reference gripper in terms of span and speed, respectively. We theoretically give the design principle, simulatively verify the method, and experimentally test this gripper on a variety of rigid, flexible, and limp objects and achieve good mechanical performance.

In order to engage in complex social interaction, humans learn at a young age to infer what others see and cannot see from a different point-of-view, and learn to predict others’ plans and behaviors. These abilities have been mostly lacking in robots, sometimes making them appear awkward and socially inept. Here we propose an end-to-end long-term visual prediction framework for robots to begin to acquire both these critical cognitive skills, known as Visual Perspective Taking (VPT) and Theory of Behavior (TOB). We demonstrate our approach in the context of visual hide-and-seek – a game that represents a cognitive milestone in human development. Unlike traditional visual predictive model that generates new frames from immediate past frames, our agent can directly predict to multiple future timestamps (25 s), extrapolating by 175% beyond the training horizon. We suggest that visual behavior modeling and perspective taking skills will play a critical role in the ability of physical robots to fully integrate into real-world multi-agent activities.

Ability to generate intelligent and generalizable facial expressions is essential for building human-like social robots. At present, progress in this field is hindered by the fact that each facial expression needs to be programmed by humans. In order to adapt robot behavior in real time to different situations that arise when interacting with human subjects, robots need to be able to train themselves without requiring human labels, as well as make fast action decisions and generalize the acquired knowledge to diverse and new contexts. We addressed this challenge by designing a physical animatronic robotic face with soft skin and by developing a vision-based self-supervised learning framework for facial mimicry. Our algorithm does not require any knowledge of the robot's kinematic model, camera calibration or predefined expression set. By decomposing the learning process into a generative model and an inverse model, our framework can be trained using a single motor babbling dataset. Comprehensive evaluations show that our method enables accurate and diverse face mimicry across diverse human subjects.

This paper presents an open-source untethered quadrupedal soft robot platform for dynamic locomotion (e.g., high-speed running and backflipping). The robot is mostly soft (80 vol.%) while driven by four geared servo motors. The robot’s soft body and soft legs were 3D printed with gyroid infill using a flexible material, enabling it to conform to the environment and passively stabilize during locomotion in multi-terrain environments. In addition, we simulated the robot in a real-time soft body simulation. With tuned gaits in simulation, the real robot can locomote at a speed of 0.9 m/s (2.5 body length/second), substantially faster than most untethered legged soft robots published to date. We hope this platform, along with its verified simulator, can catalyze agile soft robots' development.

While most robotics simulation libraries are built for low-dimensional and intrinsically serial tasks, soft-body and multi-agent robotics have created a demand for simulation environments that can model many interacting bodies in parallel. Despite the increasing interest in these fields, no existing simulation library addresses the challenge of providing a unified, highly-parallelized, GPU-accelerated interface for simulating large robotic systems. Titan is a versatile CUDA-based C++ robotics simulation library that employs a novel asynchronous computing model for GPU-accelerated simulations of robotics primitives. The innovative GPU architecture design permits simultaneous optimization and control on the CPU while the GPU runs asynchronously, enabling rapid topology optimization and reinforcement learning iterations. Kinematics are solved with a massively parallel integration scheme that incorporates constraints and environmental forces. We report dramatically improved performance over CPU-based baselines, simulating as many as 300 million primitive updates per second, while allowing flexibility for a wide range of research applications. We present several applications of Titan to high-performance simulations of soft-body and multi-agent robots.

SATNet is an award-winning MAXSAT solver that can be used to infer logical rules and integrated as a differentiable layer in a deep neural network. It had been shown to solve Sudoku puzzles visually from examples of puzzle digit images, and was heralded as an impressive achievement towards the longstanding AI goal of combining pattern recognition with logical reasoning. In this paper, we clarify SATNet's capabilities by showing that in the absence of intermediate labels that identify individual Sudoku digit images with their logical representations, SATNet completely fails at visual Sudoku (0% test accuracy). More generally, the failure can be pinpointed to its inability to learn to assign symbols to perceptual phenomena, also known as the symbol grounding problem, which has long been thought to be a prerequisite for intelligent agents to perform real-world logical reasoning. We propose an MNIST based test as an easy instance of the symbol grounding problem that can serve as a sanity check for differentiable symbolic solvers in general. Naive applications of SATNet on this test lead to performance worse than that of models without logical reasoning capabilities. We report on the causes of SATNet’s failure and how to prevent them.

Evolution provides a creative fount of complex and subtle adaptations that often surprise the scientists who discover them. However, the creativity of evolution is not limited to the natural world: Artificial organisms evolving in computational environments have also elicited surprise and wonder from the researchers studying them. The process of evolution is an algorithmic process that transcends the substrate in which it occurs. Indeed, many researchers in the field of digital evolution can provide examples of how their evolving algorithms and organisms have creatively subverted their expectations or intentions, exposed unrecognized bugs in their code, produced unexpectedly adaptations, or engaged in behaviors and outcomes, uncannily convergent with ones found in nature. Such stories routinely reveal surprise and creativity by evolution in these digital worlds, but they rarely fit into the standard scientific narrative. Instead they are often treated as mere obstacles to be overcome, rather than results that warrant study in their own right. Bugs are fixed, experiments are refocused, and one-off surprises are collapsed into a single data point. The stories themselves are traded among researchers through oral tradition, but that mode of information transmission is inefficient and prone to error and outright loss. Moreover, the fact that these stories tend to be shared only among practitioners means that many natural scientists do not realize how interesting and lifelike digital organisms are and how natural their evolution can be. To our knowledge, no collection of such anecdotes has been published before. This article is the crowd-sourced product of researchers in the fields of artificial life and evolutionary computation who have provided first-hand accounts of such cases. It thus serves as a written, fact-checked collection of scientifically important and even entertaining stories. In doing so we also present here substantial evidence that the existence and importance of evolutionary surprises extends beyond the natural world, and may indeed be a universal property of all complex evolving systems.

In this work, an automatic machine learning (AutoML) modeling architecture called Autostacker is introduced. Autostacker combines an innovative hierarchical stacking architecture and an evolutionary algorithm (EA) to perform efficient parameter search without the need for prior domain knowledge about the data or feature preprocessing. Using EA, Autostacker quickly evolves candidate pipelines with high predictive accuracy. These pipelines can be used in their given form, or serve as a starting point for further augmentation and refinement by human experts. Autostacker finds innovative machine learning model combinations and structures, rather than selecting a single model and optimizing its hyperparameters. When its performance on fifteen datasets is compared with that of other AutoML systems, Autostacker produces superior or competitive results in terms of both test accuracy and time cost.

Can machines ask questions and generate hypotheses? Despite the prevalence of big data, the process of distilling data into scientific laws has resisted automation. Particularly challenging are situations with small amounts of data that is difficult or expensive to collect, such as in robotics and other physical sciences. This talk will outline a series of recent research projects, starting with self-reflecting robotic systems, and ending with machines that can formulate hypotheses, design experiments, and interpret the results, to discover new scientific laws. We will see examples from geology to cosmology, from classical physics to modern physics, from big science to small science.

The concept of morphological computation holds that the body of an agent can, under certain circumstances, exploit the interaction with the environment to achieve useful behavior, potentially reducing the computational burden of the brain/controller. The conditions under which such phenomenon arises are, however, unclear. We hypothesize that morphological computation will be facilitated by body plans with appropriate geometric, material, and growth properties, while it will be hindered by other body plans in which one or more of these three properties is not well suited to the task. We test this by evolving the geometries and growth processes of soft robots, with either manually-set softer or stiffer material properties. Results support our hypothesis: we find that for the task investigated, evolved softer robots achieve better performances with simpler growth processes than evolved stiffer ones. We hold that the softer robots succeed because they are better able to exploit morphological computation. This four-way interaction among geometry, growth, material properties and morphological computation is but one example phenomenon that can be investigated using the system here introduced, that could enable future studies on the evolution and development of generic soft-bodied creatures.

The field of evolved virtual creatures has been suspiciously stagnant in terms of complexification of evolved agents since its inception over two decades ago. Many researchers have proposed algorithmic improvements, but none have taken hold and greatly propelled the scalability of early works. This paper suggests a more fundamental problem with co-evolving both the morphology and control of virtual creatures simultaneously one cemented in the theory of embodied cognition. We reproduce and explore in greater detail a previous finding in the literature: premature convergence of the morphology (compared to the convergence point of optimizing controllers), and discuss how this finding fits as a symptom of the proposed problem. We hope that this improved understanding of the fundamental problem domain will open the door for further scalability of evolved agents, and note that early findings from our future work point in that direction.

Difficult multiple-choice (MC) questions can be made easy by providing a set of answer options of which most are obviously wrong. In the education literature, a plethora of instructional guides exist for crafting a suitable set of wrong choices (distractors) that enable the assessment of the students' understanding. The art of MC question design thus hinges on the question-maker's experience and knowledge of the potential misconceptions. In contrast, we advocate a data-driven approach, where correct and incorrect options are assembled directly from the students' own past submissions. Large-scale online classroom settings, such as massively open online courses (MOOCs), provide an opportunity to design optimal and adaptive multiple-choice questions that are maximally informative about the students' level of understanding of the material. In this work, we (i) develop a multinomial-logit discrete choice model for the setting of MC testing, (ii) derive an optimization objective for selecting optimally discriminative option sets, (iii) propose an algorithm for finding a globally-optimal solution, and (iv) demonstrate the effectiveness of our approach via synthetic experiments and a user study. We finally showcase an application of our approach to crowd-sourcing tests from technical online forums.

Difficult test questions can be made easy by providing a set of possible answer options of which most are obviously wrong. In the education literature, a plethora of instructional guides exist for crafting a suitable set of wrong choices (distractors) in order to probe the students' understanding of the tested concept. The art of multiple-choice question design thus hinges on the question-maker's experience and knowledge of the potential misconceptions. In contrast, we advocate a data-driven approach, where correct and incorrect options are assembled directly from the students' own past submissions. Large-scale online classroom settings, such as massively open online courses (MOOCs), provide an opportunity to design optimal and adaptive multiple-choice questions that are maximally informative about the students' level of understanding of the material. We deploy a multinomial-logit discrete choice model for the setting of multiple choice testing, derive an optimization objective for selecting optimally discriminative option sets, and demonstrate the effectiveness of our approach via a user study.

This special issue displays some of the best articles presented at the Fourteenth International Conference on the Synthesis and Simulation of Living Systems (ALife 14), which was held in New York, USA, on July 30–August 2, 2014 (http://alife2014.alife.org/). ALife 14 was the fourteenth convening of ALife, one of the two largest conference series focusing on the study of artificial life, alternating with ECAL, the European edition.ALife 14 attracted a total of 203 submissions, solicited in two formats: full article (up to 8 pages) and extended abstract (2 pages). Both types were evaluated through multiple peer reviews by the Program Committee. As a result, 102 submissions were accepted for oral presentations and 62 for poster presentations. These articles and abstracts were published in the open-access conference proceedings available from MIT Press' website (https://mitpress.mit.edu/index.php?q=books/artificial-life-14). The conference program was full of scientific, professional, and creative activities, including five superb keynote talks, twenty parallel sessions covering a wide variety of topical areas, a very well-attended poster session with a plenary “Poster Blitz Movie Show,” nine workshops, six tutorials, the First Summer School of the International Society for Artificial Life (ISAL) with four lectures, a science visualization competition, and a career-advising luncheon for postdocs and graduate students. These were four vibrant, intellectually stimulating days in Manhattan.The contributions included in this special issue were selected, by the Organizing Committee and the Best Paper Committee, from the full articles presented orally at the ALife 14 conference, based on their peer review scores and the quality of their presentations. The Best Paper and the Best Student Paper of ALife 14 (http://blogs.cornell.edu/alife14nyc/best-paperposter-awards/) were also included in this special issue. Authors submitted an extended version of their conference manuscript, which went through another round of peer review and revision for journal publication. As a result, there are eight articles included in this special issue. The first four are about artificial chemistry: two discussing abstract models and two addressing real biochemistry. The last four articles are about evolution, two of them also covering morphological adaptation and computing. Their contents are summarized below.The first article, “Complex Autocatalysis in Simple Chemistries,” received the Best Paper award at ALife 14. In this article, Virgo, Ikegami, and McGregor present simple mass-action-kinetics-based artificial chemistry models and demonstrate that, even in a thermodynamically reversible chemical reaction network, complex autocatalytic cycles can emerge when certain direct reactions are prohibited. Complex nonlinear dynamics, such as bistability, arising in their simple models may offer new insight into the origins of life.In “Exploring the Space of Viable Configurations in a Model of Metabolism-Boundary Co-construction,” Agmon, Gates, Churavy, and Beer propose a spatial artificial chemistry model in which autopoietic interaction between boundary formation and metabolic reactions inside it allows stable cellular structures to emerge. They systematically examine the robustness and variability of these cellular structures by subjecting them to global or local perturbations, revealing a transition network of viable configurations.In “Computational Design of a Circular RNA with Prionlike Behavior,” Badelt, Flamm, and Hofacker present a novel theoretical chemistry method to computationally design RNA molecules that have desired properties (energy landscapes), using their software tool called the Vienna RNA package. They demonstrate the effectiveness of the proposed method by using it to design self-replicating RNA molecules that show conformational self-replication, just like the self-replication of prions.In “Compartmentalization of an all-E. coli Cell-Free Expression System for the Construction of a Minimal Cell,” Caschera and Noireaux present an experimental study on the effects of compartmentalization of their cell-free transcription-translation system extracted from E. coli. They show that, when encapsulated in cell-sized lipid vesicles, this system exhibits large fluctuations in gene expression levels, possibly due to heterogeneity in DNA template encapsulation.The next article, “An Informational Study of the Evolution of Codes and of Emerging Concepts in Populations of Agents,” was the recipient of the Best Student Paper award at ALife 14. In this article, Burgos and Polani study an agent population model in which no shared code is initially available for agents' communication and the sources of information are not distinguishable to each agent. Using evolutionary optimization, they demonstrate that a universal code can emerge if the population is well mixed, while multiple distinct codes may result in spatially structured populations. Moreover, “blind” agents can develop concepts about the environment using information coming from other agents.In “A General Statistical Method for Identifying Adaptations by Parameterizing Trait Space,” Blount proposes a formal method to detect adaptation by mapping the traits of evolving organisms to a 3D metric space that consists of variation, heritability, and differential fitness. This method does not require a priori specification of what kind of function or purpose the adaptation possesses. As a proof of concept, the calculation of differential fitness is demonstrated through the application of the proposed method to Packard's Bugs system.In “Population and Evolutionary Dynamics based on Predator-Prey Relationships in a 3D Physical Simulation,” Ito, Pilat, Suzuki, and Arita present a virtual ecosystem of 3D morphological creatures like Sims' blockies, where both ecological and evolutionary dynamics develop simultaneously and spontaneously via coevolution of predator and prey populations. Computer simulations show the emergence of short-term (ecological) and long-term (evolutionary) dynamics occurring in two different regimes of time scales, as well as their mutual interactions.Finally, in “Active Shape Discrimination with Compliant Bodies as Reservoir Computers,” Johnson, Philippides, and Husbands propose to use a randomly generated mass-spring-damper (MSD) network embedded in an agent's body as a “morphogenetic computation” device, which is a physical counterpart of neural-network-based reservoir computing. They demonstrate, through evolutionary search, that such an MSD network can be used successfully as a reservoir computer to control the agent when catching objects of a certain shape while avoiding others.We thank all the authors of the above articles for their excellent contributions, which nicely illustrate the current state of the art of artificial life. We would also like to thank the following people on the Best Paper Committee of ALife 14, who played a crucial role in evaluating numerous articles and selecting the best ones to be included in this volume:• Chair: Luis Rocha• Co-chair: Takashi Ikegami• Committee members: Joshua Auerbach, Lola Cañamero, Dominique Chu, Sylvain Cussat-Blanc, Stéphane Doncieux, Dusan Misevic, Susan Stepney, Sebastian von MammenThe success of the ALife 14 conference was due to all who helped and participated in this event, including the Program Committee members, the five keynote speakers (John H. Conway, Lee Cronin, Naomi Leonard, Jesse Louis-Rosenberg, and Karl Sims), the organizers of workshops and tutorials, local organization staff, and the following sponsors:• US National Science Foundation• MIT Press• Cornell University• Thomas J. Watson School of Engineering and Applied Science at Binghamton University• Wolfram Research• Mary Ann Liebert, Inc. PublishersFinally, we would like to give special thanks to Mark Bedau, the Editor-in-Chief of the Artificial Life journal, and Linda Reedijk, the editorial assistant, for their help, encouragement, and support in this special issue.

Soft robots have become increasingly popular in recent years -- and justifiably so. Their compliant structures and (theoretically) infinite degrees of freedom allow them to undertake tasks which would be impossible for their rigid body counterparts, such as conforming to uneven surfaces, efficiently distributing stress, and passing through small apertures. Previous work in the automated deign of soft robots has shown examples of these squishy creatures performing traditional robotic task like locomoting over flat ground. However, designing soft robots for traditional robotic tasks fails to fully utilize their unique advantages. In this work, we present the first example of a soft robot evolutionarily designed for reaching or squeezing through a small aperture -- a task naturally suited to its type of morphology. We optimize these creatures with the CPPN-NEAT evolutionary algorithm, introducing a novel implementation of the algorithm which includes multi-objective optimization while retaining its speciation feature for diversity maintenance. We show that more compliant and deformable soft robots perform more effectively at this task than their less flexible counterparts. This work serves mainly as a proof of concept, but we hope that it helps to open the door for the better matching of tasks with appropriate morphologies in robotic design in the future.

A vast majority of L1 vocabulary acquisition occurs through incidental learning during reading (Nation, 2001; Schmitt et al., 2001). We propose a probabilistic approach to generating code-mixed text as an L2 technique for increasing retention in adult lexical learning through reading. Our model that takes as input a bilingual dictionary and an English text, and generates a code-switched text that optimizes a defined “learnability” metric by constructing a factor graph over lexical mentions. Using an artificial language vocabulary, we evaluate a set of algorithms for generating code-switched text automatically by presenting it to Mechanical Turk subjects and measuring recall in a sentence completion task.

The embodied cognition paradigm emphasizes that both bodies and brains combine to produce complex behaviors, in contrast to the traditional view that the only seat of intelligence is the brain. Despite recent excitement about embodied cognition, brains and bodies remain thought of, and implemented as, two separate entities that merely interface with one another to carry out their respective roles. Previous research co-evolving bodies and brains has simulated the physics of bodies that collect sensory information and pass that information on to disembodied neural networks, which then processes that information and return motor commands. Biological animals, in contrast, produce behavior through physically embedded control structures and a complex and continuous interplay between neural and mechanical forces. In addition to the electrical pulses flowing through the physical wiring of the nervous system, the heart elegantly combines control with actuation, as the physical properties of the tissue itself (or defects therein) determine the actuation of the organ. Inspired by these phenomena from cardiac electrophysiology (the study of the electrical properties of heart tissue), we introduce electrophysiological robots, whose behavior is dictated by electrical signals flowing though the tissue cells of soft robots. Here we describe these robots and how they are evolved. Videos and images of these robots reveal lifelike behaviors despite the added challenge of having physically embedded control structures. We also provide an initial experimental investigation into the impact of different implementation decisions, such as alternatives for sensing, actuation, and locations of central pattern generators. Overall, this paper provides a first step towards removing the chasm between bodies and brains to encourage further research into physically realistic embodied cognition. Introduction and Background The fields of evolutionary robotics and artificial life have seen a great deal of emphasis on embodied cognition in recent years [Cheney et al. (2013); Bongard (2013); Rieffel et al. (2013); Auerbach and Bongard (2012); Hiller and Lipson (2012a); Lehman and Stanley (2011); Auerbach and Bongard (2010a,b); Pfeifer et al. (2007); Hornby et al. (2001); Lipson and Pollack (2000)]. There is even a paradigm called embodied cognition, which argues that the specifics of the embodiment (such as the morphology) are Figure 1: Current flowing through an evolved creature. The legend for voltage within each cell (colors) is given in Fig. 3. vital parts of the resulting behavior of the system: It argues that the co-evolutionary connection between body and brain is more deeply intertwined than the body simply acting as a minimal physical interface between the brain and the environment [Pfeifer and Bongard (2006)]. Recent work in evolutionary robotics has shown that complex behaviors can arise when co-evolving bodies and brains. At one end of the spectrum, Auerbach and Bongard (2010b) demonstrated the evolution of physical structures that had no joints or actuators, and evolved to cover the largest distance in a controlled fall due to gravity. While that work exemplifies the evolution of behavior emerging from morphology alone, it does not co-evolve any actuation or control. Auerbach and Bongard (2010a) then evolved the placement of CPG controlled rotational joints between cellular spheres, thus co-evolving morphology and control. Cheney et al. (2013) evolved locomoting soft robots made of multiple different materials: two passive voxels of differing rigidity and two actuated voxel types that expanded cyclically via out-of-phase central pattern generators (CPGs). While this work added a variety of soft materials and a new type of actuation, the pairing of muscle types directly to a CPG again reflected a focus on evolving morphology rather than sophisticated neural control. Many examples in the literature include the co-evolution of a robot morphology with an artificial neural network controller [Sims (1994); Lipson and Pollack (2000); Hornby et al. (2001); Lehman and Stanley (2011)]. These studies (and many more like them) involve what might be called “ghost” networks: artificial neural networks that provide control to the body, yet do not have any physical embodiALIFE 14: Proceedings of the Fourteenth International Conference on the Synthesis and Simulation of Living Systems Figure 2: An example of complex electrical wave propagation in cardiac modeling [Fenton et al. (2005)]. ment in the system they control. The state of input nodes to these networks is often set by sensors in the robot and output nodes typically signify behavioral outcomes in the actuators, but the computation is done supernaturally, disjoint from the body itself. In the age of 3D printing, it is a realistic goal for robots to physically walk out of a printer. It is thus worthwhile to consider designing robots that can be physically realized: i.e., those whose controllers are accounted for by being physically woven into the design of the robot. While the brains of animals are often a separate module within their bodies, animals also have central and peripheral nervous systems extending throughout their bodies. An extreme example of this is the octopus, which has as much as 90% of its neurons existing outside of its central nervous system [Zullo et al. (2009)]. The distributed and physical layout of the nervous system over space may contribute significantly to neural processing, as the delays and branching in axons (the basis for nerves) are suggested to serve computational functions [Segev and Schneidman (1999)]. Despite the prevalence of embodied, distributed circuitry in nearly all of animal life, the idea of an embodied nervous system has been absent from the field of evolutionary robotics. The sub-field called Evolvable Hardware evolves physical circuits for computer chips [Floreano and Mattiussi (2008)], but such work has not been applied to evolving the circuitry of artificial life organisms. We are unaware of work with virtual creatures that have physically embodied control systems (e.g. where neural circuitry physically runs throughout the body of the creature). We present the first such work in this paper. We propose a very basic model of electrical signal propagation throughout the body of an evolved creature. This embodied controller is based on electrophysiology (specifically at large scales, such as cardiac electrophysiology, Fig. 2). Electrophysiology is the study of the electrical properties of biological cells and tissues [Hoffman et al. (1960)]. In this model, electrical pulses from a single centralized sinusoidal pacemaker (analogous to the sinoatrial node – the pacemaker in the heart [Brown (1982)]) are propagated through the electrically conductive tissue of the creature. The location and patterning of this conductive tissue is described by an evolved Compositional Pattern Producing Network (CPPN) genome. Evolution controls the shape of the body and the electrical pathways within it, which both combine to determine the robot’s behavior. The model involves conductive tissue cells that collect voltage from neighboring cells, causing an action potential (spike) if the collected voltage exceeds the cell’s firing threshold (Fig. 3). Once this threshold is crossed, the cell depolarizes, causing a voltage spike that excites neighboring cells. This voltage spike is followed by a refractory period, during which the cell is temporarily unable to be re-excited. This model allows for the propagation of information through the body of the creature in the form of electrical signals. The structure of this flow is produced entirely by the topology of the creature and the state of each cell’s direct neighbors. In this sense, the model can be seen as a form of distributed information processing. One could draw similarities between this model and a 3D-grid of neurons, where each neuron receives inputs from, and has outputs to, its immediate neighbors. In this analogy, we are evolving where neurons should exist in the grid, what type of material the neuron is housed in, as well as the material type, if any, of grid locations that do not contain neurons. The placement of material, which is under evolutionary control, directly determines the resultant behavior of the organism. Cells that actuate will contract and expand as they depolarize (much like the contraction of cardiac muscles), leading to the locomotion behavior of the creature. In order to control the signal flow throughout the creature, insulator cells are allowed, which are unable to accept and pass on the signal. Evolution can also choose not to fill a voxel with material. The morphology of the simulated robot and tissue type at each cell is determined by a CPPN genome. This model examines the evolution of embodied cognition at a more detailed level of implementation than is typical in the literature – with embodied control circuitry resulting directly from the morphology of the individual creature. While this study only covers the classic problem of locomotion, it is a step towards truly physically embodied robots.

A long-standing, open question in biology is how populations are capable of rapidly adapting to novel environments, a trait called evolvability. A major contributor to evolvability is the fact that many biological entities are modular, especially the many biological processes and structures that can be modeled as networks, such as metabolic pathways, gene regulation, protein interactions, and animal brains. Networks are modular if they contain highly connected clusters of nodes that are sparsely connected to nodes in other clusters [4, 2]. Despite its importance and decades of research, there is no agreement on why modularity evolves [4]. Intuitively, modular systems seem more adaptable, a lesson well-known to human engineers, because it is easier to rewire a modular network with functional subunits than an entangled, monolithic network [1]. However, because this evolvability only provides a selective advantage over the long-term, such selection is at best indirect and may not be strong enough to explain the level of modularity in the natural world [4]. Modularity is likely caused by multiple forces acting to various degrees in different contexts [4], and a comprehensive understanding of the evolutionary origins of modularity involves identifying those multiple forces and their relative contributions. The leading hypothesis is that modularity mainly emerges due to rapidly changing environments that have common subproblems, but different overall problems [1]. It is unknown how much natural modularity MVG can explain, however, because it unclear if biological environments change modularly, and whether they change at a high enough frequency for this force to play a significant role. We investigate an alternate hypothesis that has been suggested, but heretofore untested, which is that modularity evolves not because it conveys evolvability, but as a byproduct from selection to reduce connection costs in a network [3].

Natural frequency tuning is a vital engineering problem. Every structure has natural frequencies, where vibrational loading at nearby frequencies excite the structure. This causes the structure to resonate, oscillating until energy is dissipated through friction or structural failure. Examples of fragility and distress from vibrational loading include civil structures during earthquakes or aircraft rotor blades. Tuning the structure's natural frequencies away from these vibrations increases the structure's robustness. Conversely, tuning towards the frequencies caused by vibrations can channel power into energy harvesting systems. Despite its importance, natural frequency tuning is often performed ad-hoc, by attaching external vibrational absorbers to a structure. This is usually adequate only for the lowest ("fundamental") resonant frequencies, yet remains standard practice due to the unintuitive and difficult nature of the problem. Given Evolutionary Algorithms' (EA's) ability to solve these types of problems, we propose to approach this problem with the EA CPPN-NEAT to evolve multi-material structures which resonate at multiple desired natural frequencies without external damping. The EA assigns the material type of each voxel within the discretized space of the object's existing topology, preserving the object's shape and using only its material composition to shape its frequency response.

In biomimetic engineering, we may take inspiration from the products of biological evolution: we may instantiate biologically realistic neural architectures and algorithms in robots, or we may construct robots with morphologies that are found in nature. Alternatively, we may take inspiration from the process of evolution: we may evolve populations of robots in simulation and then manufacture physical versions of the most interesting or more capable robots that evolve. If we follow this latter approach and evolve both the neural and morphological subsystems of machines, we can perform controlled experiments that provide unique insight into how bodies and brains can work together to produce adaptive behavior, regardless of whether such bodies and brains are instantiated in a biological or technological substrate. In this paper, we review selected projects that use such methods to investigate the synergies and tradeoffs between neural architecture, morphology, action, and adaptive behavior.

This paper presents a policy for selecting the most informative individuals in a teacher-learner type coevolution. We propose the use of the surprisal of the mean, based on Shannon information theory, which best disambiguates a collection of arbitrary and competing models based solely on their predictions. This policy is demonstrated within an iterative coevolutionary framework consisting of symbolic regression for model inference and a genetic algorithm for optimal experiment design. Complex symbolic expressions are reliably inferred using fewer than 32 observations. The policy requires 21% fewer experiments for model inference compared to the baselines and is particularly effective in the presence of noise corruption, local information content as well as high dimensional systems. Furthermore, the policy was applied in a real-world setting to model concrete compression strength, where it was able to achieve 96.1% of the passive machine learning baseline performance with only 16.6% of the data.

The connection mechanism between neighboring modules is the most critical subsystem of each module in a modular robot. Here, we describe a strong, lightweight, and solid-state connection method based on heating a low melting point alloy to form reversible soldered connections. No external manipulation is required for forming or breaking connections between adjacent connectors, making this method suitable for reconfigurable systems such as self-reconfiguring modular robots. Energy is only consumed when switching connectivity, and the ability to transfer power and signal through the connector is inherent to the method. Soldering connectors have no moving parts, are orders of magnitude lighter than other connectors, and are readily mass manufacturable. The mechanical strength of the connector is measured as 173 N, which is enough to support many robot modules, and hundreds of connection cycles are performed before failure.

Ongoing, rapid advances in three-dimensional (3D) printing technology are making it inexpensive for lay people to manufacture 3D objects. However, the lack of tools to help nontechnical users design interesting, complex objects represents a significant barrier preventing the public from benefitting from 3D printers. Previous work has shown that an evolutionary algorithm with a generative encoding based on developmental biology-a compositional pattern-producing network (CPPN)-can automate the design of interesting 3D shapes, but users collectively had to start each act of creation from a random object, making it difficult to evolve preconceived target shapes. In this paper, we describe how to modify that algorithm to allow the further evolution of any uploaded shape. The technical insight is to inject the distance to the surface of the object as an input to the CPPN. We show that this seeded-CPPN technique reproduces the original shape to an arbitrary resolution, yet enables morphing the shape in interesting, complex ways. This technology also raises the possibility of two new, important types of science: (1) It could work equally well for CPPN-encoded neural networks, meaning neural wiring diagrams from nature, such as the mouse or human connectome, could be injected into a neural network and further evolved via the CPPN encoding. (2) The technique could be generalized to recreate any CPPN phenotype, but substituting a flat CPPN representation for the rich, originally evolved one. Any evolvability extant in the original CPPN genome can be assessed by comparing the two, a project we take first steps toward in this paper. Overall, this paper introduces a method that will enable non-technical users to modify complex, existing 3D shapes and opens new types of scientific inquiry that can catalyze research on bio-inspired artificial intelligence and the evolvability benefits of generative encodings.

In 1994 Karl Sims showed that computational evolution can produce interesting morphologies that resemble natural organisms. Despite nearly two decades of work since, evolved morphologies are not obviously more complex or natural, and the field seems to have hit a complexity ceiling. One hypothesis for the lack of increased complexity is that most work, including Sims', evolves morphologies composed of rigid elements, such as solid cubes and cylinders, limiting the design space. A second hypothesis is that the encodings of previous work have been overly regular, not allowing complex regularities with variation. Here we test both hypotheses by evolving soft robots with multiple materials and a powerful generative encoding called a compositional pattern-producing network (CPPN). Robots are selected for locomotion speed. We find that CPPNs evolve faster robots than a direct encoding and that the CPPN morphologies appear more natural. We also find that locomotion performance increases as more materials are added, that diversity of form and behavior can be increased with different cost functions without stifling performance, and that organisms can be evolved at different levels of resolution. These findings suggest the ability of generative soft-voxel systems to scale towards evolving a large diversity of complex, natural, multi-material creatures. Our results suggest that future work that combines the evolution of CPPN-encoded soft, multi-material robots with modern diversity-encouraging techniques could finally enable the creation of creatures far more complex and interesting than those produced by Sims nearly twenty years ago.

The biological process of ribosomal assembly is one of the most versatile systems in nature. With only a few small building blocks, this natural process is capable of synthesizing the multitude of complex chemicals that form the basis of all organic life. This paper presents a robotics design and manufacturing scheme which seeks to capture some of the versatility of the ribosomal process. In this scheme, a custom "printer" folds a long ribbon of material in which control elements such as motors have been embedded into a morphology that is capable of accomplishing a pre-defined task. The evolved folding patterns are encoded with a special kind of compositional pattern producing network (CPPN), which can compactly describe patterns with regularities such as symmetry, repetition, and repetition with variation. This paper tests the efficacy of this design scheme and the effects of different ribbon lengths on the ability to produce walking robot morphologies. We show that a single strip of material can be folded into a variety of different morphologies displaying different forms of locomotion. Thus, the results presented here suggest a promising new method for the automated design and manufacturing of robotic systems.

The evolution of self-replication in three dimensions is explored for the first time. A discrete three-dimensional world populated with physically-realizable “molecubes” is simulated. The cubes have randomly initialized controllers, can rotate about an axis, and can attach to one another to form conglomerations. Genetic material, which defines cube controllers, is exchanged stochastically between attached cubes and subject to random mutations. Self-replicating cube conglomerations emerge in this simulation across a wide range of densities and without the use of a fitness function, yielding insight into the evolution of self-replication in nature and furthering progress toward physically-realizable self-replicating machines.

We describe a new, quadruped robot platform, Aracna, which requires non-intuitive motor commands in order to locomote and thus provides an interesting challenge for gait learning algorithms, such as those frequently developed in the Evolu-tionary Computation and Artificial Life communities. Aracna is an open-source hardware project composed of off-the-shelf and 3D-printed parts, enabling other research teams to mod-ify its design according to their scientific needs. Aracna was designed to overcome the shortcomings of a previous quadruped robot platform, whose legs were so heavy that the motors could not reliably execute the commands sent to them. We avoid this problem by locating all motors in the body core instead of on the legs and through a design which enables the servos to have a greater mechanical advantage. Specifically, each of the four legs has two joints controlled by separate four-bar linkage mechanisms that drive the pitch of the hip joint and knee joint. This novel design causes unconventional kinematics, creating an opportunity for gait-learning algorithms, which excel in counter-intuitive design spaces where human engineers tend to underperform. Be-cause it is low-cost, flexible, kinematically interesting, and and improvement over a previous design, Aracna provides a useful new hardware platform for testing algorithms that au-tomatically generate robotic behaviors.

Genetic programming has been successfully used for symbolic regression of time series data in a wide variety of applications. However, previous approaches have not taken into account the presence of multiple-time-scale dynamics despite their prevalence in both natural and artificial dynamical systems. Here, we propose an algorithm that first decomposes data from such systems into components with dynamics at different time scales and then performs symbolic regression separately for each scale. Results show that this divide-and-conquer approach improves the accuracy and efficiency with which genetic programming can be used to reverse-engineer multiple-time-scale dynamical systems.

Markerless pose inference of arbitrary subjects is a primary problem for a variety of applications, including robot vision and teaching by demonstration. Unsupervised kinematic pose inference is an ideal method for these applications as it provides a robust, training-free approach with minimal reliance on prior information. However, these methods have been considered intractable for complex models. This paper presents a general framework for inferring poses from a single depth image given an arbitrary kinematic structure without prior training. A co-evolutionary algorithm, consisting of pose and predictor populations, is applied to overcome the traditional limitations in kinematic pose inference. Evaluated on test sets of 256 synthetic and 52 real images, our algorithm shows consistent pose inference for 34 and 78 degree of freedom models with point clouds containing over 40,000 points, even in cases of significant self-occlusion. Compared to various baselines, the co-evolutionary algorithm provides at least a 3.5-fold increase in pose accuracy and a two-fold reduction in computational effort for articulated models.

Grasping has been studied from various perspectives including planning, control, and learning. In this paper, we take a learning approach to predict successful grasps for a universal jamming gripper. A jamming gripper is comprised of a flexible membrane filled with granular material, and it can quickly harden or soften to grip objects of varying shape by modulating the air pressure within the membrane. Although this gripper is easy to control, developing a physical model of its gripping mechanism is difficult because it undergoes significant deformation during use. Thus, many grasping approaches based on physical models (such as based on form- and force-closure) would be challenging to apply to a jamming gripper. Here we instead use a supervised learning algorithm and design both visual and shape features for capturing the properties of good grasps. We show that given target object data from an RGBD sensor, our algorithm can predict successful grasps for the jamming gripper without requiring a physical model. It can therefore be applied to both a parallel plate gripper and a jamming gripper without modification. We demonstrate that our learning algorithm enables both grippers to pick up a wide variety of objects, including objects from outside the training set. Through robotic experiments we are then able to define the type of objects each gripper is best suited for handling.

In systems and computational biology, much effort is devoted to functional identification of systems and networks at the molecular-or cellular scale. However, similarly important networks exist at anatomical scales such as the tendon network of human fingers: the complex array of collagen fibers that transmits and distributes muscle forces to finger joints. This network is critical to the versatility of the human hand, and its function has been debated since at least the 16(th) century. Here, we experimentally infer the structure (both topology and parameter values) of this network through sparse interrogation with force inputs. A population of models representing this structure co-evolves in simulation with a population of informative future force inputs via the predator-prey estimation-exploration algorithm. Model fitness depends on their ability to explain experimental data, while the fitness of future force inputs depends on causing maximal functional discrepancy among current models. We validate our approach by inferring two known synthetic Latex networks, and one anatomical tendon network harvested from a cadaver's middle finger. We find that functionally similar but structurally diverse models can exist within a narrow range of the training set and cross-validation errors. For the Latex networks, models with low training set error [<4%] and resembling the known network have the smallest cross-validation errors [∼5%]. The low training set [<4%] and cross validation [<7.2%] errors for models for the cadaveric specimen demonstrate what, to our knowledge, is the first experimental inference of the functional structure of complex anatomical networks. This work expands current bioinformatics inference approaches by demonstrating that sparse, yet informative interrogation of biological specimens holds significant computational advantages in accurate and efficient inference over random testing, or assuming model topology and only inferring parameters values. These findings also hold clues to both our evolutionary history and the development of versatile machines.

We describe a simple passive universal gripper, consisting of a mass of granular material encased in an elastic membrane. Using a combination of positive and negative pressure, the gripper can rapidly grip and release a wide range of objects that are typically challenging for universal grippers, such as flat objects, soft objects, or objects with complex geometries. The gripper passively conforms to the shape of a target object, then vacuum-hardens to grip it rigidly, later utilizing positive pressure to reverse this transition-releasing the object and returning to a deformable state. We describe the mechanical design and implementation of this gripper and quantify its performance in real-world testing situations. By using both positive and negative pressure, we demonstrate performance increases of up to 85% in reliability, 25% in error tolerance, and the added capability to shoot objects by fast ejection. In addition, multiple objects are gripped and placed at once while maintaining their relative distance and orientation. We conclude by comparing the performance of the proposed gripper with others in the field.

We present the automated design and manufacture of static and locomotion objects in which functionality is obtained purely by the unconstrained 3-D distribution of materials. Recent advances in multimaterial fabrication techniques enable continuous shapes to be fabricated with unprecedented fidelity unhindered by spatial constraints and homogeneous materials. We address the challenges of exploitation of the freedom of this vast new design space using evolutionary algorithms. We first show a set of cantilever beams automatically designed to deflect in arbitrary static profiles using hard and soft materials. These beams were automatically fabricated, and their physical performance was confirmed within 0.5-7.6% accuracy. We then demonstrate the automatic design of freeform soft robots for forward locomotion using soft volumetrically expanding actuator materials. One robot was fabricated automatically and assembled, and its performance was confirmed with 15% error. We suggest that this approach to design automation opens the door to leveraging the full potential of the freeform multimaterial design space to generate novel mechanisms and deformable robots.

Despite the range of applications and successes of evolutionary algorithms, expensive fitness computations often form a critical performance bottleneck. A preferred method of reducing the computational overhead is to coevolve rank predictors, providing a coarse and lightweight fitness approximation that has proven to drastically increase performance. However, the majority of previous work on rank predictor coevolution focused solely on improving the predictor heuristics while strategies to select the equally important trainer population is often an afterthought. Four different strategies are presented and benchmarked on a symbolic regression problem using hundreds of test problems with varying complexities. Of the four strategies, updating the trainer population with the solution of the highest rank variance is found to be significantly superior, resulting in a four to ten fold reduction in computational effort for similar convergence rates over the remaining strategies.

Many systems, particularly in biology and chemistry, involve the interaction of discrete quantities, such as individual elements or molecules. When the total number of elements in the system is low, the impact of individual reactions becomes non-negligible and modeling requires the simulation of exact sequences of reactions. In this paper, we introduce an algorithm that can infer an exact stochastic reaction model based on sparse measurements of an evolving system of discrete quantities. The algorithm is based on simulating a candidate model to maximize the likelihood of the data. When the likelihood is too small to provide a search gradient, the algorithm uses the distance of the data to the model's estimated distribution. Results show that this method infers stochastic models reliably with both short time gaps between measurements of the system, and long time gaps where the system state has evolved qualitatively far between each measurement. Furthermore, the proposed metric outperforms optimizing on likelihood or distance components alone. Traits measured on the search novelty, age, and bloat suggest that this algorithm scales well to increasingly complex systems.

Self-reconfiguring modular robotic systems offer a potential route to achieving programmable matter, i.e. a substance the shape and properties of which can be tuned as required to achieve a variety of tasks. However, most modular robotic system designs rely on deterministic module motions which place significant power, control, and actuation requirements on the individual modules. This leads to relatively large modules and low target structure resolution. Here we experimentally demonstrate an alternative approach based on stochastic assembly, in which modules assemble into target structures in a fluidic tank. This system employs ambient fluid motion for module transportation. Assembly is directed by controlling the fluid flow through an active assembly substrate with an array of valves. Different valving programs are used with feedback from pressure sensors to achieve the completely automated hierarchical assembly of non-planar 3D structures.

This project focuses on developing a flapping-wing hovering insect using 3D-printed wings and mechanical parts. The use of 3D printing technology has greatly expanded the possibilities for wing design, allowing wing shapes to replicate those of real insects or virtually any other shape. It has also reduced the time of a wing design cycle to a matter of minutes. An ornithopter with a mass of 3.89 g has been constructed using the 3D printing technique and has demonstrated an 85-s passively stable untethered hovering flight. This flight exhibits the functional utility of printed materials for flapping-wing experimentation and ornithopter construction and for understanding the mechanical principles underlying insect flight and control.

We explore vacuum as bonding force for modular robotics. Vacuubes are a set of modules that propagate vacuum across their interfaces in order to generate adhesive forces to form and hold structures. We use analog circuits to simulate vacuum transients and understand critical design parameters and then validate these insights in experiments. A 49-module structure that employs vacuum bonding is demonstrated. We conclude that vacuum offers a relatively strong, simple, reliable, and power-efficient connection principle.

We demonstrate the evolution of locomoting amorphous robots composed of multiple materials. Research in evolutionary robotics has traditionally been limited to morphologies comprising rigid and discrete components, such as links connected with rotational or linear joints and actuators. In the continuous robots presented here, actuation is accomplished by periodic volumetric expansion and contraction of one or more materials composing the body of the robot. The challenges of representing evolvable multi-material freeform shapes and evaluation (simulation) of the resulting soft bodies are discussed. Several genotypic representations are explored which use a level-set threshold to generate the material distribution in the phenotype. Soft body simulation of the robot is accomplished using a relaxation algorithm to model the dynamics of the resulting amorphous machines under the actuation material expansion, gravity forces, and non-linear ground friction. These results open the door to a new design space that more closely mimics the freeform, amorphous and continuous nature of biological systems.

We propose a multi-objective method for avoiding premature convergence in evolutionary algorithms, and demonstrate a three-fold performance improvement over comparable methods. Previous research has shown that partitioning an evolving population into age groups can greatly improve the ability to identify global optima and avoid converging to local optima. Here, we propose that treating age as an explicit optimization criterion can increase performance even further, with fewer algorithm implementation parameters. The proposed method evolves a population on the two-dimensional Pareto front comprising (a) how long the genotype has been in the population (age); and (b) its performance (fitness). We compare this approach with previous approaches on the Symbolic Regression problem, sweeping the problem difficulty over a range of solution complexities and number of variables. Our results indicate that the multi-objective approach identifies the exact target solution more often that the age-layered population and standard population methods. The multi-objective method also performs better on higher complexity problems and higher dimensional datasets -- finding global optima with less computational effort.

Many applications of evolutionary algorithms utilize fitness approximations, for example coarse-grained simulations in lieu of computationally intensive simulations. Here, we propose that it is better to learn approximations that accurately predict the ranks of individuals rather than explicitly estimating their real-valued fitness values. We present an algorithm that coevolves a rank-predictor which optimizes to accurately rank the evolving solution population. We compare this method with a similar algorithm that uses fitness-predictors to approximate real-valued fitnesses. We benchmark the two approaches using thousands of randomly-generated test problems in Symbolic Regression with varying difficulties. The rank prediction method showed a 5-fold reduction in computational effort for similar convergence rates. Rank prediction also produced less bloated solutions than fitness prediction.

Stochastic self-reconfiguring robots are modular robots that possess the ability to autonomously change the arrangement of their modules and do so through the use of non-deterministic processes. We present a concept for a robotic system in which the stochastic behavior of turbulent flow in a chamber is used during assembly and disassembly operations. The thermorheological properties of Pluronic®are used to implement flow routing for controlling the assembly process. This is the first use of thermorheological valving in three dimensions. A novel reversible module connection mechanism using a low melting point alloy which is soldered in a fluid environment is presented. Together with our approach to self-alignment, these are the innovations required to allow scalable self-directed assembly in three dimensions.

One of the grand challenges of self-reconfiguring modular robotics is the assembly of a functional system from thousands of components. However, to date, only systems comprised of small numbers of modules have been demonstrated. One approach to scaling to large numbers of modules is to simplify module design by relieving the modules of the typical power, control, and actuation requirements necessary for locomotion. Assembly is accomplished by taking advantage of stochastic environmental motions to move the modules into place. Here we present an experimental system in which we assemble 3D target structures stochastically from simple, 15 mm-scaled components by manipulating the fluid flow in a 1.3 L tank. We also demonstrate fundamental assembly and repair operations experimentally, and discuss initial assembly statistics.

We present a concept for a modular robot with a quasi-regular polyhedron based on a cuboctahedron element. Lattice-type modular robots can adapt their morphology by reconfiguring to various shapes. While regular polyhedrons provide the bases of many promising 3D lattice elements, few modular robots have shapes with more than six regular faces. The conceptual design and prototypes of cuboctahedron elements are presented in this paper. To account for the various connecting configurations between robotic modules, we propose a directed graph with three parameters to represent the morphology of such a modular robotic system.

We describe a robotic architecture that combines the benefits of existing enclosed robots with passive dynamics. This combination results in a mobile robot with no moving parts exposed to the environment, making it ideal for tasks where wheeled or legged robots fail. Instead of suppressing resonance, the new robot morphology relies on the dynamics of resonance for locomotion. Actuators mounted on a central sphere excite a natural mode of vibration by pulling on strings through which the sphere is mounted to a tetrahedral frame. A simple open loop controller is sufficient to cause directed motion by hopping and sliding in a prototype. Rolling as a further gait is investigated theoretically. Fully enclosed resonant dynamic robots could lead to a new type of robot locomotion powered from a vibration source only. This is useful in the microscale where traditional actuators are not available, and at the macroscale where the robot's high ruggedness is favorable.

Modular robotic systems typically assemble using deterministic processes where modules are directly placed into their target position. By contrast, stochastic modular robots take advantage of ambient environmental energy for the transportation and delivery of robot components to target locations, thus offering potential scalability. The inability to precisely predict component availability and assembly rates is a key challenge for planning in such environments. Here, we describe a computationally efficient simulator to model a modular robotic system that assembles in a stochastic fluid environment. This simulator allows us to address the challenge of planning for stochastic assembly by testing a series of potential strategies. We first calibrate the simulator using both high-fidelity computational fluid-dynamics simulations and physical experiments. We then use this simulator to study the effects of various system parameters and assembly strategies on the speed and accuracy of assembly of topologically different target structures.

Multi-material 3D-printing technologies permit the freeform fabrication of complex spatial arrangements of materials in arbitrary geometries. This technology has opened the door to a large mechanical design space with many novel yet non-intuitive possibilities. This space is not easily searched using conventional topological optimization methods such as homogenization. Here we present an evolutionary design process for three-dimensional multi-material structures that explores this design space and designs substructures tailored for custom functionalities. The algorithm is demonstrated for the design of 3D non-uniform beams and 3D compliant actuators.

A key to the success of any genetic programming process is the use of a good alphabet of atomic building blocks from which solutions can be evolved efficiently. An alphabet that is too granular may generate an unnecessarily large search space; an inappropriately coarse grained alphabet may bias or prevent finding optimal solutions. Here we introduce a method that automatically identifies a small alphabet for a problem domain. We process solutions on the complexity-optimality Pareto front of a number of sample systems and identify terms that appear significantly more frequently than merited by their size. These terms are then used as basic building blocks to solve new problems in the same problem domain. We demonstrate this process on symbolic regression for a variety of physics problems. The method discovers key terms relating to concepts such as energy and momentum. A significant performance enhancement is demonstrated when these terms are then used as basic building blocks on new physics problems. We suggest that identifying a problem-specific alphabet is key to scaling evolutionary methods to higher complexity systems.

We investigated several methods for utilizing expert knowledge in evolutionary search, and compared their impact on performance and scalability into increasingly complex problems. We collected data over one thousand randomly generated problems. We then simulated collecting expert knowledge for each problem by optimizing an approximated version of the exact solution. We then compared six different methods of seeding the approximate model in to the genetic program, such as using the entire approximate model at once or breaking it into pieces. Contrary to common intuition, we found that inserting the complete expert solution into the population is not the best way to utilize that information; using parts of that solution is often more effective. Additionally, we found that each method scaled differently based on the complexity and accuracy of the approximate solution. Inserting randomized pieces of the approximate solution into the population scaled the best into high complexity problems and was the most invariant to the accuracy of the approximate solution. Furthermore, this method produced the least bloated solutions of all methods. In general, methods that used randomized parameter coefficients scaled best with the approximate error, and methods that inserted entire approximate solutions scaled worst with the problem complexity.

In this paper we describe an optimal reconfiguration planning algorithm that morphs a grounded truss structure of known geometry into a new geometry. The plan consists of a sequence of paths to move truss elements to their new locations that generate the new truss geometry. The trusses are grounded and remain connected at all time. Intuitively, the algorithm grows gradually the new truss structure from the old one. The truss elements are rigid bars joined with 18-way connectors. The paper also introduces the design of a truss-climbing robot that can execute the plan.

Recent advancements in rapid prototyping techniques such as 3D printing and laser cutting are changing the perception of physical 3D models in architecture and industrial design. Physical models are frequently created not only to finalize a project but also to demonstrate an idea in early design stages. For such tasks, models can easily be annotated to capture comments, edits, and other forms of feedback. Unfortunately, these annotations remain in the physical world and cannot easily be transferred back to the digital world. Our system, ModelCraft, addresses this problem by augmenting the surface of a model with a traceable pattern. Any sketch drawn on the surface of the model using a digital pen is recovered as part of a digital representation. Sketches can also be interpreted as edit marks that trigger the corresponding operations on the CAD model. ModelCraft supports a wide range of operations on complex models, from editing a model to assembling multiple models, and offers physical tools to capture free-space input. Several interviews and a formal study with the potential users of our system proved the ModelCraft system useful. Our system is inexpensive, requires no tracking infrastructure or per object calibration, and we show how it could be extended seamlessly to use current 3D printing technology.

We present an algorithm that coevolves fitness predictors, optimized for the solution population, which reduce fitness evaluation cost and frequency, while maintaining evolutionary progress. Fitness predictors differ from fitness models in that they may or may not represent the objective fitness, opening opportunities to adapt selection pressures and diversify solutions. The use of coevolution addresses three fundamental challenges faced in past fitness approximation research: 1) the model learning investment; 2) the level of approximation of the model; and 3) the loss of accuracy. We discuss applications of this approach and demonstrate its impact on the symbolic regression problem. We show that coevolved predictors scale favorably with problem complexity on a series of randomly generated test problems. Finally, we present additional empirical results that demonstrate that fitness prediction can also reduce solution bloat and find solutions more reliably.

Developmental systems typically produce a phenotype through a generative process whose outcome depends on feedback from the environment. In most artificial developmental systems, this feedback occurs in one way: The environment affects the development process, but the development process does not necessarily affect the environment. Here we explore a condition where both the developing system and the environment affect each other on a similar timescale, thus resulting in system-environment dynamical interaction. Using a model inspired by termite nest construction, we demonstrate how evolution can exploit this system-environment dynamics to generate adaptive and self-repairing structure more efficiently than a purely reactive developmental system. Finally, we offer a metric to quantify the level of interaction and distinguish between reactive and interactive developmental systems.

No abstract available.

In this paper, we analyze two general-purpose encoding types, trees and graphs systematically, focusing on trends over increasingly complex problems. Tree and graph encodings are similar in application but offer distinct advantages and disadvantages in genetic programming. We describe two implementations and discuss their evolvability. We then compare performance using symbolic regression on hundreds of random nonlinear target functions of both 1-dimensional and 8-dimensional cases. Results show the graph encoding has less bias for bloating solutions but is slower to converge and deleterious crossovers are more frequent. The graph encoding however is found to have computational benefits, suggesting it to be an advantageous trade-off between regression performance and computational effort.

In this paper we propose a genetic programming approach to learning stochastic models with unsymmetrical noise distributions. Most learning algorithms try to learn from noisy data by modeling the maximum likelihood output or least squared error, assuming that noise effects average out. While this process works well for data with symmetrical noise distributions (such as Gaussian observation noise), many real-life sources of noise are not symmetrically distributed, thus this approach does not hold. We suggest improved learning can be obtained by including noise sources explicitly in the model as a stochastic element. A stochastic element is a random sub-process or latent variable of a hidden system that can propagate nonlinear noise to the observable outputs. Stochastic elements can skew and distort output features making regression of analytical models particularly difficult and error minimizing approaches inhibiting. We introduce a new method to infer the analytical model of a system by decomposing non-uniform noise observed at the outputs into uniform stochastic elements appearing symbolically inside the system. Results demonstrate the ability to regress exact analytical models where stochastic elements are embedded inside nonlinear and polynomial hidden systems.

Long-term physical survivability of most robotic systems today is achieved through durable hardware. In contrast, most biological systems are not made of robust materials; long-term sustainability and evolutionary adaptation in nature are provided through processes of self-repair and, ultimately, self-reproduction. Here we demonstrate a large space of possible robots capable of autonomous self-reproduction. These robots are composed of actuated modules equipped with electromagnets to selectively control the morphology of the robotic assembly. We show a variety of 2-D and 3-D machines from 3 to 2n modules, and two physical implementations that each achieves two generations of reproduction. We show both automatically generated and manually designed morphologies

A major challenge in interactive evolution is extracting user preferences with minimal probing. We introduce an interactive multi-objective coevolutionary algorithm that actively selects the most informative probes: We simultaneously coevolve a population of candidate models that explain users' selection so far, and a population of candidate probes that cause the most divergence among model predictions, thereby elucidating model uncertainties (divergence). As progress is made, we begin selecting for probes with the highest expected outcome averaged among different models, thereby exploiting model certainties (consensus). In the evolution of pen stroke drawings, we find this technique to be highly effective at extracting preference models from very limited human interaction. Using only pair-wise preference questions, strategy and preference in pen stroke drawings are extracted in fewer than ten user probes. Our results show that the optimal questions to probe the user need not include drawings similar to the target drawing. Instead, the user models converge on trends in the user responses, thereby extrapolating strong preference for target drawings which the models are never actually trained to prefer.

With the availability of affordable new desktop fabrication techniques such as 3D printing and laser cutting, physical models are used increasingly often during the architectural and industrial design cycle. Models can easily be annotated to capture comments, edits and other forms of feedback. Unfortunately, these annotations remain in the physical world and cannot be easily transferred back to the digital world. Here we present a simple solution to this problem based on a tracking pattern printed on the surface of each model. Our solution is inexpensive, requires no tracking infrastructure or per object calibration, and can be used in the field without a computer nearby. It lets users not only capture annotations, but also edit the model using a simple yet versatile command system. Once captured, annotations and edits are merged into the original CAD models. There they can be easily edited or further refined. We present the design of a SolidWorks plug-in implementing this concept, and report initial feedback from potential users using our prototype. We also present how this prototype could be extended seamlessly to a fully functional system using current 3D printing technology.

The static properties of tensegrity structures have been widely appreciated in civil engineering as the basis of extremely lightweight yet strong mechanical structures. However, the dynamic properties and their potential utility in the design of robots have been relatively unexplored. This paper introduces robots based on tensegrity structures, which demonstrate that the dynamics of such structures can be utilized for locomotion. Two tensegrity robots are presented: TR3, based on a triangular tensegrity prism with three struts, and TR4, based on a quadrilateral tensegrity prism with four struts. For each of these robots, simulation models are designed, and automatic design of controllers for forward locomotion are performed in simulation using evolutionary algorithms. The evolved controllers are shown to be able to produce static and dynamic gaits in both robots. A real-world tensegrity robot is then developed based on one of the simulation models as a proof of concept. The results demonstrate that tensegrity structures can provide the basis for lightweight, strong, and fault-tolerant robots with a potential for a variety of locomotor gaits

In previous papers we have described a co-evolutionary algorithm (EEA), the estimation-exploration algorithm, that infers the hidden inner structure of systems using minimal testing. In this paper we introduce the concept of 'managed challenge' to alleviate the problem of disengagement in this and other co-evol-utionary algorithms. A known problem in co-evolutionary dynamics occurs when one population systematically outperforms the other, resulting in a loss of selection pressure for both populations. In system identification (which deals with determining the inner structure of a system using only input/output data), multiple trials (a test that causes the system to produce some output) on the system to be identified must be performed. When such trials are costly, this disengagement results in wasted data that is not utilized by the evolutionary process. Here we propose that data from futile interactions should be stored during disengagement and automatically re-introduced later, when the population re-engages: we refer to this as the test bank. We demonstrate that the advantage of the test bank is two-fold: it allows for the discovery of more accurate models, and it reduces the amount of required training data for both parametric identification -- parameterizing inner structure -- and symbolic identification -- approximating inner structure using symbolic equations -- of nonlinear systems.

Tensegrity structures are stable 3-dimensional mechanical structures which maintain their form due to an intricate balance of forces between disjoint rigid elements and continuous tensile elements. Tensegrity structures can give rise to lightweight structures with high strength-to-weight ratios and their utility has been appreciated in architecture, engineering and recently robotics. However, the determination of connectivity patterns of the rigid and tensile elements which lead to stable tensegrity is challenging. Available methods are limited to the use of heuristic guidelines, hierarchical design based on known components, or mathematical methods which can explore only a subset of the space. This paper investigates the use of evolutionary algorithms in the form-finding of tensegrity structures. It is shown that an evolutionary algorithm can be used to explore the space of arbitrary tensegrity structures which are difficult to design using other methods, and determine new, non-regular forms. It suggests that evolutionary algorithms can be used as the basis for a general design methodology for tensegrity structures.

This paper investigates the hypothesis that robots based on highly coupled mechanical structures can give rise to redundancy in control. Highly coupled mechanical structures have the property that actuation at one location can translate into movement at multiple locations, and conversely, movement at one location can be caused by multiple actuators. Due to this property, multiple control strategies may exist for a single behavior. Tensegrity structures which have recently been shown to form the basis for successful locomotor robots (Paul et al., 2005), have highly coupled mechanical structures. Thus, as a case study, it was of interest to investigate whether these new tensegrity based robots could offer a high degree of redundancy of control. This was investigated on two robots, based on three and four strut tensegrity prisms. Control strategies for locomotion were evolved using a genetic algorithm in simulation, and the evolved behaviors were compared. It was found that multiple control strategies existed for forward locomotion in both structures, and that qualitatively similar behavior could be obtained with significantly different control strategies. This indicated that a considerable degree of redundancy could exist in the control of robots based on highly coupled mechanical structures.

This paper describes an active learning approach to the problem of grammatical inference, specifically the inference of deterministic finite automata (DFAs). We refer to the algorithm as the estimation-exploration algorithm (EEA). This approach differs from previous passive and active learning approaches to grammatical inference in that training data is actively proposed by the algorithm, rather than passively receiving training data from some external teacher. Here we show that this algorithm outperforms one version of the most powerful set of algorithms for grammatical inference, evidence driven state merging (EDSM), on randomly-generated DFAs. The performance increase is due to the fact that the EDSM algorithm only works well for DFAs with specific balances (percentage of positive labelings), while the EEA is more consistent over a wider range of balances. Based on this finding we propose a more general method for generating DFAs to be used in the development of future grammatical inference algorithms.

We present a coevolutionary algorithm for inferring the topology and parameters of a wide range of hidden nonlinear systems with a minimum of experimentation on the target system. The algorithm synthesizes an explicit model directly from the observed data produced by intelligently generated tests. The algorithm is composed of two coevolving populations. One population evolves candidate models that estimate the structure of the hidden system. The second population evolves informative tests that either extract new information from the hidden system or elicit desirable behavior from it. The fitness of candidate models is their ability to explain behavior of the target system observed in response to all tests carried out so far; the fitness of candidate tests is their ability to make the models disagree in their predictions. We demonstrate the generality of this estimation-exploration algorithm by applying it to four different problems-grammar induction, gene network inference, evolutionary robotics, and robot damage recovery-and discuss how it overcomes several of the pathologies commonly found in other coevolutionary algorithms. We show that the algorithm is able to successfully infer and/or manipulate highly nonlinear hidden systems using very few tests, and that the benefit of this approach increases as the hidden systems possess more degrees of freedom, or become more biased or unobservable. The algorithm provides a systematic method for posing synthesis or analysis tasks to a coevolutionary system.

Remote robotics applications, such as space exploration or operation in hazardous environments, would greatly benefit from automated recovery algorithms for unanticipated failure or damage. In this paper a two-stage evolutionary algorithm is introduced-which we call the estimation-exploration algorithm-that forwards this aim by first evolving a damage hypothesis after failure and then re-evolving a compensatory neural controller to restore functionality. The algorithm presupposes that a robot simulator is running continuously onboard the physical robot. In this paper, the 'physical' robot is also simulated, but in future work the algorithm will be applied to a real, physical robot. Although evolutionary algorithms require a large number of evaluations to produce a useful solution, the results reported here indicate that almost complete functionality can be restored after only three evaluations on the 'physical' robot, as opposed to over 3000 evaluations if the compensatory controller is evolved all on the 'physical' robot. Our algorithm also has the benefit of producing a diagnostic model of the failure.

Implementations of self-reconfigurable robotics rearrange modules through a planned, deterministic reconfiguration path. Reconfiguration is achieved using active module locomotion or manipulation. Here we propose a form of self-reconfigurable robotics based on passive, stochastic self-organization. Solid-state cellular units exploit "Brownian motion" in their environment and require no local power or locomotion ability. This form of reconfiguration avoids many of the barriers that prevent self-reconfigurable robotics from extending to large numbers and small scales. We demonstrate working prototypes and discuss preliminary analytical and computational models for analyzing the scalability of this concept.

The field of evolutionary robotics has demonstrated the ability to automatically design the morphology and controller of simple physical robots through synthetic evolutionary processes. However, it is not clear if variation-based search processes can attain the complexity of design necessary for practical engineering of robots. Here, we demonstrate an automatic design system that produces complex robots by exploiting the principles of regularity, modularity, hierarchy, and reuse. These techniques are already established principles of scaling in engineering design and have been observed in nature, but have not been broadly used in artificial evolution. We gain these advantages through the use of a generative representation, which combines a programmatic representation with an algorithmic process that compiles the representation into a detailed construction plan. This approach is shown to have two benefits: it can reuse components in regular and hierarchical ways, providing a systematic way to create more complex modules from simpler ones; and the evolved representations can capture intrinsic properties of the design space, so that variations in the representations move through the design space more effectively than equivalent-sized changes in a nongenerative representation. Using this system, we demonstrate for the first time the evolution and construction of modular, three-dimensional, physically locomoting robots, comprising many more components than previous work on body-brain evolution.

Recent research has demonstrated the ability for automatic design of the morphology and control of real physical robots using techniques inspired by biological evolution. The main criticism of the evolutionary design approach, however, is that it is doubtful whether it will reach the high complexities necessary for practical engineering. Here we claim that for automatic design systems to scale in complexity the designs they produce must be made of re-used modules. Our approach is based on the use of a generative design grammar subject to an evolutionary process. Unlike a direct encoding of a design, a generative design specification can re-use components, giving it the ability to create more complex modules from simpler ones. Re-used modules are also valuable for improved efficiency in testing and construction. We describe a system for creating generative specifications capable of hierarchical modularity by combining Lindenmayer systems with evolutionary algorithms. Using this system we demonstrate for the first time a generative system for physical, modular, 2D locomoting robots and their controllers.

October 01 2001 Evolutionary Robotics: The Biology, Intelligence and Technology of Self-Organizing Machines Evolutionary Robotics: The Biology, Intelligence and Technology of Self-Organizing Machines. StefanoNolfi and DarioFloreano. ( 2000, MIT Press). $50.00 hardcover, 320 pages. Hod Lipson Hod Lipson Cornell Computer Aided Design Lab, Mechanical and Aerospace Engineering and Computing and Information Science, Cornell University, Ithaca, NY 14853-2801 USA, [email protected] Search for other works by this author on: This Site Google Scholar Author and Article Information Hod Lipson Cornell Computer Aided Design Lab, Mechanical and Aerospace Engineering and Computing and Information Science, Cornell University, Ithaca, NY 14853-2801 USA, [email protected] Online Issn: 1530-9185 Print Issn: 1064-5462 © 2002 Massachusetts Institute of Technology2002 Artificial Life (2001) 7 (4): 419–424. https://doi.org/10.1162/106454601317297031 Cite Icon Cite Permissions Share Icon Share Twitter LinkedIn Views Icon Views Article contents Figures & tables Video Audio Supplementary Data Peer Review Search Site Citation Hod Lipson; Evolutionary Robotics: The Biology, Intelligence and Technology of Self-Organizing Machines. Artif Life 2001; 7 (4): 419–424. doi: https://doi.org/10.1162/106454601317297031 Download citation file: Ris (Zotero) Reference Manager EasyBib Bookends Mendeley Papers EndNote RefWorks BibTex toolbar search Search Dropdown Menu nav search search input Search input auto suggest search filter All ContentAll JournalsArtificial Life Search Advanced Search This content is only available as a PDF. © 2002 Massachusetts Institute of Technology2002 Article PDF first page preview Close Modal You do not currently have access to this content.

The difficulties associated with designing, building, and controlling robots have led their development to a stasis: Applications are limited mostly to repetitive tasks with predefined behavior. Over the last few years we have been trying to address this challenge through an alternative approach: Rather than trying to control an existing machine or create a general-purpose robot, we propose that both the morphology and the controller should evolve at the same time. This process can lead to the automatic design of special-purpose mechanisms and controllers for specific short-term objectives. Here we provide a brief review of three generations of our recent research, which underlies the robots shown on the cover of this issue: Automatically designed static structures, automatically designed and manufactured dynamic electromechanical systems, and modular robots automatically designed through a generative DNA-like encoding.

We propose a new process for continuously reconfigurable robotics. Given a task, a robot evolves a suitable morphology and control, then prints the assembled 3D structure, downloads into it and performs the task, and then recycles into a different form for the next task. This approach relinquishes the need to adhere to discrete or fixed components as well as manually designed configurations. Although the technology for fully realizing the proposed concept is not entirely available, we have demonstrated a first case: Given the task of locomotion, various robots with different mechanics and control are evolved automatically. The design space is comprised of only linear actuators and sigmoidal control neurons embodied in an arbitrary thermoplastic body. The robots then print pre-assembled using rapid prototyping technology, and perform the task in reality. The robots are then recycled. This paper described one implementation and provides examples of successful physical robots generated by the proposed process.

This article introduces a method for clustering irregularly shaped data arrangements using high-order neurons. Complex analytical shapes are modeled by replacing the classic synaptic weight of the neuron by high-order tensors in homogeneous coordinates. In the first- and second-order cases, this neuron corresponds to a classic neuron and to an ellipsoidalmetric neuron. We show how high-order shapes can be formulated to follow the maximum-correlation activation principle and permit simple local Hebbian learning. We also demonstrate decomposition of spatial arrangements of data clusters, including very close and partially overlapping clusters, which are difficult to distinguish using classic neurons. Superior results are obtained for the Iris data.

An important key to reconstructing a three-dimensional object depicted by a two-dimensional line drawing projection is face identification. Identification of edge circuits in a 2D projection corresponding to actual faces of a 3D object becomes complex when the projected object is in wireframe representation. This representation is commonly encountered in drawings made during the conceptual design stage of mechanical parts. When nonmanifold objects are considered, the situation becomes even more complex. This paper discusses the principles underlying face identification and presents an algorithm capable of performing this identification. Face-edge-vertex relationships applicable to nonmanifold objects are also proposed. Examples from a working implementation are given.