Robots that can adapt like animals
Open accessThe DOI matches its Crossref record; 39 of the 40 references it cites were found in a registry or an index.
Abstract
Robots have transformed many industries, most notably manufacturing, and have the power to deliver tremendous benefits to society, such as in search and rescue, disaster response, health care and transportation. They are also invaluable tools for scientific exploration in environments inaccessible to humans, from distant planets to deep oceans. A major obstacle to their widespread adoption in more complex environments outside factories is their fragility. Whereas animals can quickly adapt to injuries, current robots cannot 'think outside the box' to find a compensatory behaviour when they are damaged: they are limited to their pre-specified self-sensing abilities, can diagnose only anticipated failure modes, and require a pre-programmed contingency plan for every type of potential damage, an impracticality for complex robots. A promising approach to reducing robot fragility involves having robots learn appropriate behaviours in response to damage, but current techniques are slow even with small, constrained search spaces. Here we introduce an intelligent trial-and-error algorithm that allows robots to adapt to damage in less than two minutes in large search spaces without requiring self-diagnosis or pre-specified contingency plans. Before the robot is deployed, it uses a novel technique to create a detailed map of the space of high-performing behaviours. This map represents the robot's prior knowledge about what behaviours it can perform and their value. When the robot is damaged, it uses this prior knowledge to guide a trial-and-error learning algorithm that conducts intelligent experiments to rapidly discover a behaviour that compensates for the damage. Experiments reveal successful adaptations for a legged robot injured in five different ways, including damaged, broken, and missing legs, and for a robotic arm with joints broken in 14 different ways. This new algorithm will enable more robust, effective, autonomous robots, and may shed light on the principles that animals use to adapt to injury.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Claims other papers make about this one
Each is checked against this paper’s own text: does it say what the citing paper says it does?
The cited source doesn’t support the claim as written.
In the citing paper
“We render images using the MuJoCo Physics Engine's [45] built-in renderer. This renderer is not intended to be photo-realistic, and physically plausible choices of textures and lighting are not needed.”
From Domain randomization for transferring deep neural networks from simulation to the real world (2017) · cites this paper as [45]
Checked by a reasoning model — may be wrong.
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“under Quality-Diversity optimization (Pugh et al., 2016), the search space is divided into behavioral niches to evolve diverse solutions, a paradigm that challenged the dominant approach of training a single agent, as it proved more robust in real-world settings (Cully et al., 2015).”
From Dynamics of niche construction in adaptable populations evolving in diverse environments (2023) · cites this paper as Cully et al., 2015
Experimental — accuracy not yet measuredThe source backs part of the claim, not all of it — or backs it for a different population, condition or comparison.
In the citing paper
“Similarly, extensive reliance on simulation and offline training further distances learning from the full variability and constraint structure of real sensorimotor interaction (Jakobi et al., 1995; Cully et al., 2015).”
From Toward Enactive Artificial Intelligence (2026) · cites this paper as Cully et al., 2015
Experimental — accuracy not yet measured
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Gaussian Processes for Machine Learning
Rasmussen, C. E., & Williams, C. K. I. (2005) · doi:10.7551/mitpress/3206.001.0001 · Cited by 11,705
Corke, P., Roberts, J., Cunningham, J., & Hainsworth, D. (2008) · Springer Handbook of Robotics · doi:10.1007/978-3-540-30301-5_50 · Cited by 3,543
A survey of robot learning from demonstration
Argall, B. D., Chernova, S., Veloso, M., & Browning, B. (2009) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2008.10.024 · Cited by 3,227
Reinforcement learning in robotics: A survey
Kober, J., Bagnell, J. A., & Peters, J. (2013) · The International Journal of Robotics Research · doi:10.1177/0278364913495721 · Cited by 3,160 · cited 9×
“The key insight is that, whereas current learning algorithms either start with no knowledge of the search space 17 or with minimal knowledge from a few human demonstrations 17,18 , animals better understand the space of possible behaviors and their value from previous experience 19 , enabling…”
BackgroundMethodResultInfluentialDiagnosis and Fault-Tolerant Control
Blanke, M., Kinnaert, M., Lunze, J., & Staroswiecki, M. (2003) · Springer eBooks · doi:10.1007/978-3-662-05344-7 · Cited by 2,789
Nonlinear Programming: Sequential Unconstrained Minimization Techniques
Fiacco, A. V., & McCormick, G. P. (1968) · Defense Technical Information Center (DTIC) · Cited by 2,262
Stanley: The robot that won the DARPA Grand Challenge
Thrun, S., Montemerlo, M., Dahlkamp, H., Stavens, D., Aron, A., Diebel, J., Fong, P., Gale, J., Halpenny, M., Hoffmann, G., et al. (2006) · Journal of Field Robotics · doi:10.1002/rob.20147 · Cited by 2,237 · cited 2×
“3C, undam-aged: 24 [16; 41] seconds, 3 [2; 5] physical trials, repaired: 29 [16; 82] seconds, 3.5 [2; 10] trials).”
BackgroundInfluentialBrochu, E., Cora, V. M., & de Freitas, N. (2010) · arXiv (Cornell University) · doi:10.48550/arxiv.1012.2599 · Cited by 2,011
Bayesian integration in sensorimotor learning
Körding, K. P., & Wolpert, D. M. (2004) · Nature · doi:10.1038/nature02169 · Cited by 2,004 · cited 1×
“…optimization procedure followed by our robot appears similar to the technique employed by humans when they optimize an unknown function 23 , and there is strong evidence that animal brains learn probability distributions, combine them with prior knowledge, and act as Bayesian optimizers 25,26 .”
MethodControl of mental activities by internal models in the cerebellum
Ito, M. (2008) · Nature Reviews Neuroscience · doi:10.1038/nrn2332 · Cited by 1,356 · cited 1×
“A final parallel is that the simulator and Gaussian process components of Intelligent Trial and Error are two forms of predictive models, which are known to exist in animals 29,12 .”
Research works printed but not linked: 2
- Siciliano, B. & Khatib, O. Springer Handbook of Robotics (Springer, 2008). · doi:10.1007/978-3-540-30301-5
- Kluger, J. & Lovell, J. Apollo 13 (Mariner Books, 2006).
Cited by
Deep neural networks are easily fooled: High confidence predictions for unrecognizable images
Nguyen, A., Yosinski, J., & Clune, J. (2015) · 2015 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2015.7298640 · Cited by 3,258
Self-citationDomain randomization for transferring deep neural networks from simulation to the real world
Tobin, J., Fong, R., Ray, A., Schneider, J., Zaremba, W., & Abbeel, P. (2017) · 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8202133 · Cited by 3,124
“[8] demonstrates that exploration in the real world can be more efficient with prior knowledge from a simulator.”
BackgroundAdversarial Examples: Attacks and Defenses for Deep Learning
Yuan, X., He, P., Zhu, Q., & Li, X. (2019) · IEEE Transactions on Neural Networks and Learning Systems · doi:10.1109/tnnls.2018.2886017 · Cited by 1,733
“Then in each iteration, MAP-Elites, like general EA algorithm, chose a random organism, mutated them randomly, and replaced with the current ones if the new ones have higher fitness (high certainty for a class of a neural network).”
MethodInfluentialRahwan, I., Cebrian, M., Obradovich, N., Bongard, J., Bonnefon, J.-F., Breazeal, C., Crandall, J. W., Christakis, N. A., Couzin, I. D., Jackson, M. O., et al. (2019) · Nature · doi:10.1038/s41586-019-1138-y · Cited by 1,364
Machine learning in materials science
Wei, J., Chu, X., Sun, X., Xu, K., Deng, H., Chen, J., Wei, Z., & Lei, M. (2019) · InfoMat · doi:10.1002/inf2.12028 · Cited by 1,069
von Rueden, L., Mayer, S., Beckh, K., Georgiev, B., Giesselbach, S., Heese, R., Kirsch, B., Walczak, M., Pfrommer, J., Pick, A., et al. (2021) · IEEE Transactions on Knowledge and Data Engineering · doi:10.1109/tkde.2021.3079836 · Cited by 793
“For instance, the walking traits of bipedal and hexapod robots can be optimized based on locomotion controller simulation [16], [73], or the stability of stacked objects can be simulated with physics engines [72], [74].”
BackgroundMethodInfluentialSchneider, G. (2018) · Nature Reviews Drug Discovery · doi:10.1038/nrd.2017.232 · Cited by 910
Such, F. P., Madhavan, V., Conti, E., Lehman, J., Stanley, K. O., & Clune, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1712.06567 · Cited by 525
“being high performing), seeking to collect a set of highperforming, yet interestingly different policies (Mouret & Clune, 2015; Lehman & Stanley, 2011b; Cully et al., 2015; Pugh et al., 2016).”
BackgroundMethodSelf-citationDesigning neural networks through neuroevolution
Stanley, K. O., Clune, J., Lehman, J., & Miikkulainen, R. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-018-0006-z · Cited by 689
Self-citationSim-to-Real: Learning Agile Locomotion For Quadruped Robots
Tan, J., Zhang, T., Coumans, E., Iscen, A., Bai, Y., Hafner, D., Bohez, S., & Vanhoucke, V. (2018) · Robotics: Science and Systems XIV · doi:10.15607/rss.2018.xiv.010 · Cited by 735
Sim-to-Real: Learning Agile Locomotion For Quadruped Robots
Tan, J., Zhang, T., Coumans, E., Iscen, A., Bai, Y., Hafner, D., Bohez, S., & Vanhoucke, V. (2018) · arXiv · doi:10.48550/arxiv.1804.10332 · Cited by 734
A light-driven artificial flytrap
Wani, O. M., Zeng, H., & Priimagi, A. (2017) · Nature Communications · doi:10.1038/ncomms15546 · Cited by 703
Solving Rubik's Cube with a Robot Hand
OpenAI, Akkaya, I., Andrychowicz, M., Chociej, M., Litwin, M., McGrew, B., Petron, A., Paino, A., Plappert, M., Powell, G., et al. (2019) · arXiv · doi:10.48550/arxiv.1910.07113 · Cited by 602
“This is in contrast to prior work that explicitly searched over a set of policies until it found one that works for a broken robot [22].”
ResultBessa, M. A., Bostanabad, R., Liu, Z., Hu, A., Apley, D. W., Brinson, C., Chen, W., & Liu, W. K. (2017) · Computer Methods in Applied Mechanics and Engineering · doi:10.1016/j.cma.2017.03.037 · Cited by 573
“…protein analysis and genetics [77]; biotechnology with the design of new drugs [78,79]; artificial intelligence with robots adapting like animals [80] and programs mastering the GO game [81]; psychology by profiling neuropsychosocial behavior [82]; climate change and its effects on food [83];…”
BackgroundBayesian Optimization for Adaptive Experimental Design: A Review
Greenhill, S., Rana, S., Gupta, S., Vellanki, P., & Venkatesh, S. (2020) · IEEE Access · doi:10.1109/access.2020.2966228 · Cited by 520
“Recent years have seen new applications in areas such as robotics [45], [46],”
BackgroundQuality Diversity: A New Frontier for Evolutionary Computation
Pugh, J. K., Soros, L. B., & Stanley, K. O. (2016) · Frontiers in Robotics and AI · doi:10.3389/frobt.2016.00040 · Cited by 514
“(2015) introduce an alternate QD algorithm called multi-dimensional archive of phenotypic elites (MAP-Elites) that collects elite versions of diverse behavior within individual bins in a behavioral map.”
BackgroundMethodInfluentialThe current state and future outlook of rescue robotics
Delmerico, J., Mintchev, S., Giusti, A., Gromov, B., Melo, K., Horvat, T., Cadena, C., Hutter, M., Ijspeert, A., Floreano, D., et al. (2019) · Journal of Field Robotics · doi:10.1002/rob.21887 · Cited by 371
Ecoffet, A., Huizinga, J., Lehman, J., Stanley, K. O., & Clune, J. (2021) · Nature · doi:10.1038/s41586-020-03157-9 · Cited by 330
Self-citationA review on self-healing polymers for soft robotics
Terryn, S., Langenbach, J., Roels, E., Brancart, J., Bakkali-Hassani, C., Poutrel, Q.-A., Georgopoulou, A., George Thuruthel, T., Safaei, A., Ferrentino, P., et al. (2021) · Materials Today · doi:10.1016/j.mattod.2021.01.009 · Cited by 309
Attentional Correlation Filter Network for Adaptive Visual Tracking
Choi, J., Chang, H. J., Yun, S., Fischer, T., Demiris, Y., & Choi, J. Y. (2017) · 2017 IEEE Conference on Computer Vision and Pattern Recognition (CVPR) · doi:10.1109/cvpr.2017.513 · Cited by 311
“A method to predict the performance of thousands of different robotic behaviours based on a pre-computed performance map was presented in [4].”
MethodWang, R., Lehman, J., Clune, J., & Stanley, K. O. (2019) · arXiv
“effectively co-evolve, novelty search with local competition [33], MAP-Elites [34, 35], Innovation Engines [36], which exploit the opportunity to transfer high-quality solutions from one objective among many to another, and the combinatorial multi-objective evolutionary algorithm (CMOEA), which extends the Innovation…”
BackgroundSelf-citationOsman, A. I., Nasr, M., Eltaweil, A. S., Hosny, M., Farghali, M., Al-Fatesh, A. S., Rooney, D. W., & Abd El-Monaem, E. M. (2024) · International Journal of Hydrogen Energy · doi:10.1016/j.ijhydene.2024.03.223 · Cited by 270
Conti, E., Madhavan, V., Such, F. P., Lehman, J., Stanley, K. O., & Clune, J. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1712.06560 · Cited by 108
“That is the approach of novelty search (Lehman & Stanley, 2011a) and quality diversity algorithms (Cully et al., 2015; Mouret & Clune, 2015; Pugh et al., 2016), which are described in detail below.”
BackgroundMethodInfluentialSelf-citationA literature survey of the robotic technologies during the COVID-19 pandemic
Wang, X. V., & Wang, L. (2021) · Journal of Manufacturing Systems · doi:10.1016/j.jmsy.2021.02.005 · Cited by 246
“Self-engineering systems, like self-adapting robots, were predicted to be helpful in the complex working environment [57].”
Programmable gear-based mechanical metamaterials
Fang, X., Wen, J., Cheng, L., Yu, D., Zhang, H., & Gumbsch, P. (2022) · Nature Materials · doi:10.1038/s41563-022-01269-3 · Cited by 246
Quality and Diversity Optimization: A Unifying Modular Framework
Cully, A., & Demiris, Y. (2018) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2017.2704781 · Cited by 258
Self-citationRobot learning towards smart robotic manufacturing: A review
Liu, Z., Liu, Q., Xu, W., Wang, L., & Zhou, Z. (2022) · Robotics and Computer-Integrated Manufacturing · doi:10.1016/j.rcim.2022.102360 · Cited by 215
Generalization guides human exploration in vast decision spaces
Wu, C. M., Schulz, E., Speekenbrink, M., Nelson, J. D., & Meder, B. (2018) · Nature Human Behaviour · doi:10.1038/s41562-018-0467-4 · Cited by 242
A soft robot that adapts to environments through shape change
Shah, D. S., Powers, J. P., Tilton, L. G., Kriegman, S., Bongard, J., & Kramer-Bottiglio, R. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-00263-1 · Cited by 202
“Although many simulation to reality (“sim2real”) methods have been reported [35, 36, 24, 37, 25, 38, 39, 40, 41, 42, 43], none have documented the transfer from simulation to reality of shape-changing robots.”
BackgroundMulti-expert learning of adaptive legged locomotion
Yang, C., Yuan, K., Zhu, Q., Yu, W., & Li, Z. (2020) · Science Robotics · doi:10.1126/scirobotics.abb2174 · Cited by 201
“The contact sequence—such as the contact location, timing, and duration—needs to be specified either by manual design or by an additional planner (21, 22).”
BackgroundIntrinsically Motivated Goal Exploration Processes with Automatic Curriculum Learning
Forestier, S., Portelas, R., Mollard, Y., & Oudeyer, P. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1708.02190 · Cited by 166
Robotic metamorphosis by origami exoskeletons
Miyashita, S., Guitron, S., Li, S., & Rus, D. (2017) · Science Robotics · doi:10.1126/scirobotics.aao4369 · Cited by 208
“The least squares problem (10) can be solved using the Moore-Penrose pseudo inverse of the matrix J”
Go-Explore: a New Approach for Hard-Exploration Problems
Ecoffet, A., Huizinga, J., Lehman, J., Stanley, K. O., & Clune, J. (2019) · arXiv · doi:10.48550/arxiv.1901.10995 · Cited by 205
“The idea of preserving and exploring from stepping stones in an archive comes from the quality diversity (QD) family of algorithms (like MAP-elites [60, 79] and novelty search with local competition [80]), and Go-Explore is an enhanced QD algorithm based on MAP-Elites.”
BackgroundMethodSelf-citationVassiliades, V., Chatzilygeroudis, K., & Mouret, J.-B. (2018) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2017.2735550 · Cited by 151
“For instance, it has been used to produce: behavioral repertoires that enable robots to adapt to damage in a matter of minutes [25], [26], perform complex tasks [27], or even adapt to damage while completing their”
BackgroundMethodInfluentialSelf-citationEntrepreneurial action, creativity, & judgment in the age of artificial intelligence
Townsend, D. M., & Hunt, R. A. (2019) · Journal of Business Venturing Insights · doi:10.1016/j.jbvi.2019.e00126 · Cited by 174
High-dimensional Bayesian optimization using low-dimensional feature spaces
Moriconi, R., Deisenroth, M. P., & Sesh Kumar, K. S. (2020) · Machine Learning · doi:10.1007/s10994-020-05899-z · Cited by 175
“Neural networks as an explicit feature map within an MGP have already been applied successfully for modeling non-smooth responses in robot locomotion [9,11].”
MethodThe Rise of Artificial Intelligence under the Lens of Sustainability
Khakurel, J., Penzenstadler, B., Porras, J., Knutas, A., & Zhang, W. (2018) · Technologies · doi:10.3390/technologies6040100 · Cited by 187
“(e.g., robots that can adapt like animals [23], game designed by Google “Alphago” [24]) Similarly, Huang et al. [18] discussed four types of intelligences.”
BackgroundEvolution-Guided Policy Gradient in Reinforcement LearningPreprint
Khadka, S., & Tumer, K. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1805.07917 · Cited by 300
“Some examples include incorporating crossover and better mutation operators [21], adaptive exploration noise [20, 40], and explicit diversity maintenance techniques [8, 9, 29, 52].”
BackgroundMethodAntBot: A six-legged walking robot able to home like desert ants in outdoor environments
Dupeyroux, J., Serres, J. R., & Viollet, S. (2019) · Science Robotics · doi:10.1126/scirobotics.aau0307 · Cited by 179
“These advances, in addition to recent findings on walking recovery in legged robots that have been damaged during the foraging trip (61), open up many possibilities for legged robots in the field of real-world navigation.”
BackgroundNti, E. K., Cobbina, S. J., Attafuah, E. E., Opoku, E., & Gyan, M. A. (2022) · Sustainable Futures · doi:10.1016/j.sftr.2022.100068 · Cited by 195
Lehman, J., Clune, J., Misevic, D., Adami, C., Altenberg, L., Beaulieu, J., Bentley, P. J., Bernard, S., Beslon, G., Bryson, D. M., et al. (2020) · Artificial Life · doi:10.1162/artl_a_00319 · Cited by 181
“Yet both philosophical arguments [23, 26, 27, 82, 109] and high-profile publications [2, 16, 18 – 21, 52, 59, 60, 83, 85, 86, 92, 141, 146, 148] suggest that digital evolution can be a useful tool to aid and complement the study of biological evolution.”
Self-citationSimulation Intelligence: Towards a New Generation of Scientific MethodsPreprint
Lavin, A., Krakauer, D. C., Zenil, H., Gottschlich, J., Mattson, T., Brehmer, J., Anandkumar, A., Choudry, S., Rocki, K., Baydin, A. G., et al. (2021) · CaltechAUTHORS (California Institute of Technology) · doi:10.48550/arxiv.2112.03235 · Cited by 221
Evolving embodied intelligence from materials to machines
Howard, D., Eiben, A. E., Kennedy, D. F., Mouret, J.-B., Valencia, P., & Winkler, D. (2019) · Nature Machine Intelligence · doi:10.1038/s42256-018-0009-9 · Cited by 161
Self-citationLearning to Walk in the Real World with Minimal Human EffortPreprint
Ha, S., Xu, P., Tan, Z., Levine, S., & Tan, J. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2002.08550 · Cited by 261
“Note that our approach is different from repertoire-based adaptation [10, 11] that relies on the pre-obtained gait library.”
MethodSolving Multitask Optimization Problems With Adaptive Knowledge Transfer via Anomaly Detection
Wang, C., Liu, J., Wu, K., & Wu, Z. (2021) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2021.3068157 · Cited by 170
A Survey on Policy Search Algorithms for Learning Robot Controllers in a Handful of Trials
Chatzilygeroudis, K., Vassiliades, V., Stulp, F., Calinon, S., & Mouret, J.-B. (2020) · IEEE Transactions on Robotics · doi:10.1109/tro.2019.2958211 · Cited by 160
“For instance, some experiments assume that demonstrations can be provided, but that they are imperfect [13], [14]; some others assume that a damaged robot knows its model in its intact form, but not the damaged model [15]–[17].”
BackgroundMethodInfluentialSelf-citationShi, Z., Yang, W., Deng, X., Cai, C., Yan, Y., Liang, H., Liu, Z., & Qiao, Z. (2020) · Molecular Systems Design & Engineering · doi:10.1039/d0me00005a · Cited by 158
Adaptive and Resilient Soft Tensegrity Robots
Rieffel, J. A., & Mouret, J.-B. (2018) · Soft Robotics · doi:10.1089/soro.2017.0066 · Cited by 164
“Cully et al. (Cully et al., 2015) also used a non-constant prior; however, (i) they generated it using a physics simulator, which is especially challenging for a vibrating tensegrity robot, and (ii) they only computed this prior for a discrete set of potential solutions, which, in turn, constrain…”
InfluentialSelf-citationNeuroevolution in Games: State of the Art and Open Challenges
Risi, S., & Togelius, J. (2015) · IEEE Transactions on Computational Intelligence and AI in Games · doi:10.1109/tciaig.2015.2494596 · Cited by 173
Increasing generality in machine learning through procedural content generation
Risi, S., & Togelius, J. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-0208-z · Cited by 122
“The Map-Elites algorithm in particular, originally introduced to create more robust robot gaits [59], has been adapted to create sets of game levels that vary in what skills they require from the agent or what mechanics they feature [60].”
BackgroundKnowledge-integrated machine learning for materials: lessons from gameplaying and robotics
Hippalgaonkar, K., Li, Q., Wang, X., Fisher, J. S., Kirkpatrick, J., & Buonassisi, T. (2023) · Nature Reviews Materials · doi:10.1038/s41578-022-00513-1 · Cited by 174
“…approaches that combine prior domain knowledge with active learning, to tackle increasingly complex problems spanning several workflow steps The goal of Stage 3 is to improve the quality and speed of outcomes, and has been an increasing focus of the robotics 184,185 and gameplaying 186 communities.”
BackgroundEmergence of robust self-organized undulatory swimming based on local hydrodynamic force sensing
Thandiackal, R., Melo, K., Paez, L., Herault, J., Kano, T., Akiyama, K., Boyer, F., Ryczko, D., Ishiguro, A., & Ijspeert, A. J. (2021) · Science Robotics · doi:10.1126/scirobotics.abf6354 · Cited by 144
“In addition, the modular design of the robot with its redundant components for pattern generation based on central and/or peripheral control provides a mechanism for high fault tolerance (81, 82).”
BackgroundLiquid metals and their hybrids as stimulus–responsive smart materials
Ren, L., Xu, X., Du, Y., Kalantar‐Zadeh, K., & Dou, S. X. (2019) · Materials Today · doi:10.1016/j.mattod.2019.10.007 · Cited by 150
Covariance matrix adaptation for the rapid illumination of behavior space
Fontaine, M. C., Togelius, J., Nikolaidis, S., & Hoover, A. K. (2020) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3377930.3390232 · Cited by 210
“robot behaviors [10, 13, 14], without explicitly searching for each behavior itself. A particularly useful application is generating diverse locomotive gaits so a robot can quickly respond to damage [11]. QD can automatically generate many interesting video game levels by diversifying content in response to how an AI…”
BackgroundMethodInfluentialEvolution of Collective Behaviors for a Real Swarm of Aquatic Surface Robots
Duarte, M., Costa, V., Gomes, J., Rodrigues, T., Silva, F., Oliveira, S. M., & Christensen, A. L. (2016) · PLOS ONE · doi:10.1371/journal.pone.0151834 · Cited by 134
“We are considering two different approaches to address these issues in future work: (i) further improve the simulated movement dynamics, thereby decreasing the mismatch between simulation and reality, and (ii) select for controllers that rely on motion patterns that transfer well from simulation to reality, as propose…”
BackgroundMethodRobot Learning From Randomized Simulations: A Review
Muratore, F., Ramos, F., Turk, G., Yu, W., Gienger, M., & Peters, J. (2022) · Frontiers in Robotics and AI · doi:10.3389/frobt.2022.799893 · Cited by 128
“The work by Cully et al. (2015) can be seen as both static and adaptive domain randomization, where a large set of hexapod locomotion policies is learned before execution on the physical robot, and subsequently evaluated in simulation.”
BackgroundAdaptive Morphology: A Design Principle for Multimodal and Multifunctional Robots
Mintchev, S., & Floreano, D. (2016) · IEEE Robotics & Automation Magazine · doi:10.1109/mra.2016.2580593 · Cited by 131
“[101], adaptive controllers can compensate for morphological failures in less then two minutes.”
Panda, M. K., Centore, R., Causà, M., Tuzi, A., Borbone, F., & Naumov, P. (2016) · Scientific Reports · doi:10.1038/srep29610 · Cited by 119
“The images of protiated IMACET (denoted H6 for clarity; green circles), IMACET-D2 (D2, yellow circles) and IMACET-D6 (D6, purple circles) in panels (d–g) were extracted from high-speed video recordings (recording rate: (2-5)·103 s−1).”
Adaptive Control Strategies for Interlimb Coordination in Legged Robots: A Review
Aoi, S., Manoonpong, P., Ambe, Y., Matsuno, F., & Wörgötter, F. (2017) · Frontiers in Neurorobotics · doi:10.3389/fnbot.2017.00039 · Cited by 136
“Besides these bio-inspired control approaches, Cully et al. (2015) proposed an alternative machine learning based approach consisting of two main parts: an automatically generated, precomputed, behavior-performance map, and a trial-and-error learning algorithm (Figure 7).”
BackgroundMethodInfluentialThe evolutionary origins of hierarchy
Mengistu, H., Huizinga, J., Mouret, J.-B., & Clune, J. (2016) · PLOS Computational Biology · doi:10.1371/journal.pcbi.1004829 · Cited by 115
“We performed this search with the multi-dimensional archive of phenotypic elites (MAP-Elites) algorithm [56, 57].”
MethodSelf-citationEmbodied Evolution in Collective Robotics: A Review
Bredeche, N., Haasdijk, E., & Prieto, A. (2018) · Frontiers in Robotics and AI · doi:10.3389/frobt.2018.00012 · Cited by 103
“More recently, Bongard et al. (2006) and Cully et al. (2015) addressed resilience by introducing fast online re-optimization to recover from hardware damage.”
Reset-free Trial-and-Error Learning for Robot Damage Recovery
Chatzilygeroudis, K. I., Vassiliades, V., & Mouret, J.-B. (2017) · Robotics and Autonomous Systems · doi:10.1016/j.robot.2017.11.010 · Cited by 135
“An alternative approach is to let a damaged robot learn a new policy by trial-and-error, which bypasses the need for a diagnosis [8, 7, 9].”
BackgroundMethodInfluentialSelf-citation15 challenges for AI: Or what AI (currently) can’t do
Hagendorff, T., & Wezel, K. (2020) · AI & SOCIETY · doi:10.1007/s00146-019-00886-y · Cited by 113
“(Lehman et al. 2018), for example such as the surprising behavior of robots in computer simulations (Cully et al. 2015).”
BackgroundSim-to-Real Transfer for Biped Locomotion
Yu, W., Kumar, V. C., Turk, G., & Liu, C. K. (2019) · 2019 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros40897.2019.8968053 · Cited by 101
“The closest work to ours is that of Cully et al. [31].”
BackgroundEvolutionary Reinforcement Learning: A Survey
Bai, H., Cheng, R., & Jin, Y. (2023) · Intelligent Computing · doi:10.34133/icomputing.0025 · Cited by 137
“, walking strategies) to help a robot adapt quickly to various kinds of damages [18], and has achieved better performance and robustness than PPO for simulated hexapod robot tasks [181].”
BackgroundScalable co-optimization of morphology and control in embodied machines
Cheney, N., Bongard, J., SunSpiral, V., & Lipson, H. (2018) · Journal of The Royal Society Interface · doi:10.1098/rsif.2017.0937 · Cited by 151
“While recent successes have demonstrated the potential of effective optimization for the control policies of agents with fixed morphologies [17], [13], [27] or – to a lesser extent – the optimization of morphologies (body plans) for agents with minimal and fixed control policies [12], [10], [1], the co-optimization of…”
BackgroundNolfi, S. (2022) · The MIT Press eBooks · doi:10.7551/mitpress/13780.003.0008 · Cited by 111
The ecology and evolution of autotomy
Emberts, Z., Escalante, I., & Bateman, P. W. (2019) · Biological reviews/Biological reviews of the Cambridge Philosophical Society · doi:10.1111/brv.12539 · Cited by 109
“Moreover, engineers have started incorporating similar biomimetic and bio-inspired designs for limb loss compensation into their research (Cully et al., 2015).”
BackgroundMeta Reinforcement Learning with Latent Variable Gaussian Processes
Sæmundsson, S., Hofmann, K., & Deisenroth, M. P. (2018) · arXiv · doi:10.48550/arxiv.1803.07551 · Cited by 106
“Although meta learning given multiple or non-stationary dynamics has also been considered in [11, 18, 4, 1].”
BackgroundOpen Issues in Evolutionary Robotics
Silva, F., Duarte, M., Correia, L., Oliveira, S. M., & Christensen, A. L. (2016) · Evolutionary Computation · doi:10.1162/evco_a_00172 · Cited by 106
“…a variety of different, high-performing behaviors throughout evolution instead of a single general behavior, and the transferability-based behavioral repertoire algorithm (Cully and Mouret, 2015), which combines NSLC with the transferability approach (Koos et al., 2013) discussed in Section 2.1.”
BackgroundMethodInfluentialSelf-reconfigurable multilegged robot swarms collectively accomplish challenging terradynamic tasks
Ozkan-Aydin, Y., & Goldman, D. I. (2021) · Science Robotics · doi:10.1126/scirobotics.abf1628 · Cited by 121
Skills for physical artificial intelligence
Miriyev, A., & Kovač, M. (2020) · Nature Machine Intelligence · doi:10.1038/s42256-020-00258-y · Cited by 111
Quality-Diversity Optimization: A Novel Branch of Stochastic Optimization
Chatzilygeroudis, K. I., Cully, A., Vassiliades, V., & Mouret, J.-B. (2021) · Springer optimization and its applications · doi:10.1007/978-3-030-66515-9_4 · Cited by 170
Self-citationArtificial Intelligence in Elite Sports—A Narrative Review of Success Stories and Challenges
Hammes, F., Hagg, A., Asteroth, A., & Link, D. (2022) · Frontiers in Sports and Active Living · doi:10.3389/fspor.2022.861466 · Cited by 118
“Quality diversity (QD) algorithms search for diversity of good solutions (Cully et al., 2015).”
BackgroundMethodAutonomous skill discovery with quality-diversity and unsupervised descriptors
Cully, A. (2019) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3321707.3321804 · Cited by 130
“On one hand, QD algorithms have shown to be instrumental in generating large collections of solutions [5, 28], oen leading to surprisingly creative results [4, 17].”
BackgroundSelf-citationBlack-box data-efficient policy search for robotics
Chatzilygeroudis, K., Rama, R., Kaushik, R., Goepp, D., Vassiliades, V., & Mouret, J.-B. (2017) · 2017 IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2017.8202137 · Cited by 94
Self-citationReal-time emotion generation in human-robot dialogue using large language models
Mishra, C., Verdonschot, R. G., Peter Hagoort, A., & Skantze, G. (2023) · Frontiers in Robotics and AI · doi:10.3389/frobt.2023.1271610 · Cited by 93
Strain-controlled power devices as inspired by human reflex
Zhang, S., Ma, B., Zhou, X., Hua, Q., Gong, J. Q., Liu, T., Cui, X. Y., Zhu, J., Guo, W., Jing, L., et al. (2020) · Nature Communications · doi:10.1038/s41467-019-14234-7 · Cited by 94
Learning Fast Adaptation With Meta Strategy Optimization
Yu, W., Tan, J., Bai, Y., Coumans, E., & Ha, S. (2020) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2020.2974685 · Cited by 87
Bioinspired electronics for intelligent soft robots
Bang, J., Choi, S. H., Pyun, K. R., Jung, Y., Hong, S., Kim, D., Lee, Y., Won, D., Jeong, S., Shin, W., et al. (2024) · Nature Reviews Electrical Engineering · doi:10.1038/s44287-024-00081-2 · Cited by 96
“Future system should integrate sensitive performance metrics capable of detecting damage, actuation mechanisms to avoid harmful stimuli and learning algorithms that enable the robot to adapt to and compensate for its damaged state while maintaining functionality 223 .”
Policy gradient assisted MAP-Elites
Nilsson, O., & Cully, A. (2021) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3449639.3459304 · Cited by 144
“Typically the number of optimized parameters is kept below 100 [13, 15, 40].”
InfluentialSelf-citationThe blame shift: Robot service failures hold service firms more accountable
Ryoo, Y., Jeon, Y. A., & Kim, W. (2024) · Journal of Business Research · doi:10.1016/j.jbusres.2023.114360 · Cited by 89
Policy search in continuous action domains: An overview
Sigaud, O., & Stulp, F. (2019) · Neural Networks · doi:10.1016/j.neunet.2019.01.011 · Cited by 102
Discovering the elite hypervolume by leveraging interspecies correlation
Vassiliades, V., & Mouret, J.-B. (2018) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3205455.3205602 · Cited by 108
“The purpose of this experiment is to create a repertoire of joint angles for a redundant robotic arm for which the resulting end effector positions cover its reachable space [10].”
BackgroundMethodInfluentialSelf-citationLiquid Metal Smart Materials toward Soft Robotics
Chen, S., Wang, H., Liu, T.-Y., & Liu, J. (2023) · Advanced Intelligent Systems · doi:10.1002/aisy.202200375 · Cited by 86
Recent Advances in Perceptive Intelligence for Soft Robotics
Lin, Z., Wang, Z., Zhao, W., Xu, Y., Wang, X., Zhang, T., Sun, Z., Lin, L., & Peng, Z. (2023) · Advanced Intelligent Systems · doi:10.1002/aisy.202200329 · Cited by 90
RETRACTED ARTICLE: A review on quantum computing and deep learning algorithms and their applications
Valdez, F., & Melín, P. (2022) · Soft Computing · doi:10.1007/s00500-022-07037-4 · Cited by 92
RetractedCorucci, F., Cheney, N., Giorgio-Serchi, F., Bongard, J., & Laschi, C. (2018) · Soft Robotics · doi:10.1089/soro.2017.0055 · Cited by 105
“More-over, increasingly sophisticated techniques that preserve the effectiveness of simulated design once deployed in the real world are being developed, 11,12 which will allow a more direct transposition from simulation to reality.”
Scalable sim-to-real transfer of soft robot designs
Kriegman, S., Nasab, A. M., Shah, D., Steele, H., Branin, G., Levin, M. G., Bongard, J., & Kramer-Bottiglio, R. K. (2020) · IEEE International Conference on Soft Robotics (RoboSoft) · doi:10.1109/robosoft48309.2020.9116004 · Cited by 100
“[5] to neatly visualize the predicted fitness of a very large library of control policies, as a function of the time a robot’s six limbs were in contact with the simulated ground plane: 6D quinary control space was mapped to 2D, by nesting pairs of dimensions within each other.”
BackgroundInsect-Inspired Robots: Bridging Biological and Artificial Systems
Manoonpong, P., Patané, L., Xiong, X., Brodoline, I., Dupeyroux, J., Viollet, S., Arena, P., & Serres, J. (2021) · Sensors · doi:10.3390/s21227609 · Cited by 83
“[219] proposed an intelligent trial-and-error algorithm that allows a hexapod robot to automatically learn a behavior map consisting of over 13,000 high-performing behaviors or gaits.”
BackgroundMethodGrounded Action Transformation for Robot Learning in Simulation
Hanna, J., & Stone, P. (2017) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v31i1.11044 · Cited by 72
“Cully et al. (2015) use a simulator to estimate fitness values for low-dimensional robot behaviors which gives the robot prior knowledge of how to adapt its behavior if it becomes damaged during real world operation.”
MethodDiversity policy gradient for sample efficient quality-diversity optimization
Pierrot, T., Macé, V., Chalumeau, F., Flajolet, A., Cideron, G., Beguir, K., Cully, A., Sigaud, O., & Perrin, N. (2022) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3512290.3528845 · Cited by 106
“The principal advantage of QD approaches resides in their intrinsic capacity to deliver a large and diverse set of working alternatives when a single solution fails (Cully et al., 2015).”
BackgroundInfluentialSelf-citationTerryn, S., Brancart, J., Roels, E., Verhelle, R. R., Safaei, A., Cuvellier, A., Vanderborght, B., & Assche, G. V. (2022) · Macromolecules · doi:10.1021/acs.macromol.2c00434 · Cited by 82
Surrogate Modeling for Bayesian Optimization Beyond a Single Gaussian Process
Lu, Q., Polyzos, K. D., Li, B., & Giannakis, G. B. (2023) · IEEE Transactions on Pattern Analysis and Machine Intelligence · doi:10.1109/tpami.2023.3264741 · Cited by 78
“ANUMBER of machine learning and artificial intelligence (AI) applications boil down to optimizing an ‘expensive-to-evaluate’ black-box function, including hyperparameter tuning [1], drug discovery [2], and policy optimization in robotics [3].”
Prompt to Transfer: Sim-to-Real Transfer for Traffic Signal Control with Prompt Learning
Da, L., Gao, M., Mei, H., & Wei, H. (2024) · Proceedings of the AAAI Conference on Artificial Intelligence · doi:10.1609/aaai.v38i1.27758 · Cited by 119
“Unlike system identification approaches (Cutler, Walsh, and How 2014; Cully et al. 2015) that try to learn the precise physical parameters, Grounded Action Trans-formation (GAT) (Hanna and Stone 2017) does not require a parameterized simulator that can be modified.”
MethodNever Stop Learning: The Effectiveness of Fine-Tuning in Robotic Reinforcement LearningPreprint
Julian, R., Swanson, B., Sukhatme, G. S., Levine, S., Finn, C., & Hausman, K. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.10190 · Cited by 102
“…et al., 2019), via online search of a pre-learned representation of the space models, policies, or high-level skills (Chatzilygeroudis et al., 2018; Cully et al., 2015; Kaushik et al., 2020; Merel et al., 2019), or by learning physics simulation parameters from real data (Rastogi et al.; Jeong et…”
BackgroundData-Efficient Domain Randomization With Bayesian Optimization
Muratore, F., Eilers, C., Gienger, M., & Peters, J. (2021) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2021.3052391 · Cited by 77
Data-efficient Domain Randomization with Bayesian Optimization
Muratore, F., Eilers, C., Gienger, M., & Peters, J. (2020) · arXiv · doi:10.48550/arxiv.2003.02471 · Cited by 77
Quality diversity for multi-task optimization
Mouret, J.-B., & Maguire, G. (2020) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3377930.3390203 · Cited by 95
“MAP-Elites [25] is the most used archive-based algorithm because it is conceptually simple and leads to good results [4, 7, 11, 14, 16, 22].”
BackgroundMethodInfluentialSelf-citationTransactive Energy for Flexible Prosumers Using Algorithmic Game Theory
Τσαούσογλου, Γ., Pinson, P., & Paterakis, N. G. (2021) · IEEE Transactions on Sustainable Energy · doi:10.1109/tste.2021.3055764 · Cited by 76
Metadata
- Authors
- Cully, A., Clune, J., Tarapore, D., & Mouret, J.-B.
- Affiliations
- Centre National de la Recherche Scientifique (FR)
- University of Wyoming (US)
- Corresponding author
- Mouret, Jean-Baptiste
- Venue
- Nature
- Published
- 2015-05-27
- Volume
- 521
- Issue
- 7553
- Pages
- 503-507
- ISSN
- 0028-0836 (ISSN-L)
- 1476-4687
- OpenAlex
- W1738827650
- Cited by
- 979
- References
- 40
- Citations by year
- 2026: 51 · 2025: 76 · 2024: 88 · 2023: 90 · 2022: 100 · 2021: 115 · 2020: 129 · 2019: 112 · 2018: 78 · 2017: 78 · 2016: 48 · 2015: 13
- FWCI
- 70.47
- Type
- journal article
- Language
- en
- Open access
- green
- arxiv.org
- Licence
- www.springer.com/tdm (tdm)
- www.springer.com/tdm (vor)
- Locations
- Nature (publishedVersion)
- PubMed (publishedVersion)
- ePrints Soton (University of Southampton) (acceptedVersion)
- arXiv (Cornell University) (submittedVersion)
- HAL (submittedVersion, other-oa)
- RePEc: Research Papers in Economics (submittedVersion)
- arXiv (Cornell University)
- Funders
- Agence Nationale de la Recherche (ANR-12-JS03-0009)
- Topics
- Physical Sciences › Computer Science › Artificial Intelligence › Reinforcement Learning in Robotics
- Physical Sciences › Engineering › Mechanical Engineering › Modular Robots and Swarm Intelligence
- Physical Sciences › Computer Science › Artificial Intelligence › Evolutionary Algorithms and Applications
- Keywords
- compensatory behavior, trial-and-error learning, autonomous mobile robots, legged robots, robot manipulators, prior knowledge, robot robustness, search and rescue robots, disaster response
- MeSH terms
- Adaptation, Physiological, Algorithms, Animals, Artificial Intelligence, Behavior, Animal, Dogs, Extremities, Motor Skills, Robotics, Time Factors, Biomimetics
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.