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
Items 801–900 of 1,007
Fault Recovery Through Online Adaptation of Boolean Network Robots
Baldini, P., Braccini, M., & Roli, A. (2025) · Sensors · doi:10.3390/s25185849 · Cited by 3
“Differently, a robot displaying fault recovery could overcome issues in movement actuation by decreasing the speed of the movement or by changing its strategy altogether [8–10].”
Sirithunge, C., Chandiramani, V., Xie, Y. H., Häuser, H., Conn, A. T., & Fumiya Iida, A. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11247489 · Cited by 1
“Next generation robots are expected to be self-sustained in controller adaptation, morphological adaptation [21], [22] and learning [23], physically resilient to damage [24] and capable of integrating with collective systems [25].”
θ l θ u -Parametric Multitask Optimization: Joint Search in Solution and Infinite Task Spaces
Wei, T., Liu, J., Gupta, A., Tan, P. S., & Ong, Y.-S. (2025) · IEEE Transactions on Evolutionary Computation · doi:10.1109/tevc.2025.3598333 · Cited by 7
“Taking robotic control as an example [11], [12], standard MTO can be configured for multiple predefined morphologies of robots to achieve specific tasks including navigation, manipulation, and locomotion.”
Iteratively Learn Diverse Strategies with State Distance Information
Fu, W., Du, W., Li, J., Chen, S., Zhang, J., & Wu, Y. (2023) · neural information processing systems · doi:10.52202/075280-0991 · Cited by 4
Agile legged locomotion in reconfigurable modular robots
Yu, C., Matthews, D., Wang, J., Gu, J., Blackiston, D. J., Rubenstein, M., & Kriegman, S. (2026) · Proceedings of the National Academy of Sciences · doi:10.1073/pnas.2519129123 · Cited by 8
“Compared to other legged robots that adapt to damage but contain a single point of failure (37, 38), the modular bodies described above have a greater capacity to survive large unforeseen changes to their structure because their structure contains self - sufficient submachines, any subset of which…”
Qin, H., Xiang, Y., Han, Y., Wang, Y., Wu, C., & Feng, F. (2026) · IEEE Transactions on Emerging Topics in Computational Intelligence · doi:10.1109/tetci.2026.3670689 · Cited by 1
“TheapproachusesthewidelyadoptedMAP-Elites framework[49],whichhasdemonstratedstrongperformance acrossrobotics[50],[51],gaming[52],[53],andsoftware testing[54],[55].”
Rojas, L., Vega, E., Jorquera, L., & García, J. (2026) · Mathematics · doi:10.3390/math14071091 · Cited by 2
Zhang, H., Zhou, F., Xue, Y., & Xue, Y. (2026) · Sensors · doi:10.3390/s26123790 · Cited by 1
“The BO [10,20] rapidly locked the search range into the low-x 1 region.”
Damage Adaptation in Seconds for Architected Materials
Avtges, J., Ketchum, J., Young, H., Kim, T., Truby, R. L., & Murphey, T. D. (2026) · Robotics: Science and Systems · doi:10.15607/rss.2026.xxii.158 · Cited by 1
TLQD at the ICST 2026 Tool Competition – UAV Testing Track
Zhang, H., Zhou, S., Qin, H., & Xiang, Y. (2026) · IEEE International Conference on Software Testing, Verification and Validation (ICST) · doi:10.1109/icst69053.2026.00074 · Cited by 1
“To address these challenges, this paper first formulates the UAV test case generation problem as a quality–diversity (QD) optimization model [6], and then designs a UAV test case generation tool based on a two-level quality–diversity approach, which is implemented on the Aerialist platform.”
Learning Human Search Strategies From a Crowdsourcing Game
Sexton, R., & Ren, M. Y. (2016) · ASME International Design Engineering Technical Conferences and Computers and Information in Engineering Conference (IDETC/CIE) · doi:10.1115/detc2016-59775 · Cited by 1
Dettmann, A., Born, A., Bartsch, S., & Kirchner, F. (2015) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros.2015.7354017 · Cited by 1
“Since these actions are usually robot-independent, the reactive action layer needs to decompose each action into 1The author is with the Faculty of Mathematics and Computer Science, University of Bremen, 28359 Bremen, Germany firstname.lastname@uni-bremen.de 2The author is with the German Research…”
BackgroundSwitching Adaptability in Human-Inspired Sidesteps: A Minimal Model
Fujii, K., Yoshihara, Y., Tanabe, H., & Yamamoto, Y. (2017) · Frontiers in Human Neuroscience · doi:10.3389/fnhum.2017.00298 · Cited by 1
“Current technology can succeed in reproducing such real-time adaptation in video game tasks (Mnih et al., 2015) and overcoming unpredicted deficits (Yoshihara et al., 2007; Cully et al., 2015), although such adaptation is limited to a certain part of the agent’s body.”
BackgroundAnimal-guided evolutionary computation in honeybees and robots
Mariano, P., Salem, Z., Mills, R., Zahadat, P., Correia, L., & Schmickl, T. (2017) · European Conference on Artificial Life · doi:10.1162/isal_a_085 · Cited by 1
AI personalities: clues from animal research
Winter, G. D. (2018) · Journal of Experimental & Theoretical Artificial Intelligence · doi:10.1080/0952813x.2018.1430861 · Cited by 1
Free Simulation Software and Library
Uğurlu, B., & Ivaldi, S. (2017) · Springer eBooks · doi:10.1007/978-94-007-7194-9_27-1 · Cited by 1
Policy transfer methods in RoboCup keep-away
Didi, S., & Nitschke, G. (2018) · Genetic and Evolutionary Computation Conference Companion (GECCO Companion) · doi:10.1145/3205651.3205710 · Cited by 1
“Non-objective search methods [14] such as novelty search [12], out-perform objective based search in various control tasks defined by complex, high dimensional and deceptive fitness landscapes [3], [10].”
BackgroundA Learning Based Recovery for Damaged Snake-Like Robots
Guan, Z., Huang, J., Jian, Z., liu, L., Cheng, L., & Huang, K. (2018) · Lecture notes in computer science · doi:10.1007/978-3-030-04239-4_3 · Cited by 1
Objective versus Non-Objective Search in Evolving Morphologically Robust Robot Controllers
Putter, R., & Nitschke, G. (2018) · IEEE Symposium Series on Computational Intelligence (SSCI) · doi:10.1109/ssci.2018.8628627 · Cited by 2
“Recent work in evolutionary controller design for ER systems indicated that maintaining phenotypic (behavioral) diversity in populations of potential solutions improves the quality (task performance) of evolved robot behaviors [11], [12], [13], [7].”
BackgroundInfluentialToward faster reinforcement learning for robotics applications by using Gaussian processes
Younes, A., & Yushchenko, A. S. (2019) · AIP conference proceedings · doi:10.1063/1.5133351 · Cited by 1
Influences of Neural Network Structures on an Efficient Reinforcement Learning Policy Search
Zhu, W., & Rosendo, A. (2019) · IEEE International Conference on Robotics and Biomimetics (ROBIO) · doi:10.1109/robio49542.2019.8961475 · Cited by 1
“A few learningbased algorithms [1], [2] or model-based algorithms [3] can help robots discover the environment, adapt their trajectories”
BackgroundResultRoveda, L., Forgione, M., & Piga, D. (2020) · International Conference on Ubiquitous Robots (UR) · doi:10.1109/ur49135.2020.9144761 · Cited by 1
“In particular, [25] employs constrained BO to select three force-controller parameters for combined position/interaction tasks; [26] describes a trial-and-error algorithm that allows robots to adapt their behaviour in presence of damage; [27] proposes a methodology to achieve automatic tuning of…”
MethodDiversity-based design assist for large legged robots
Howard, G., Löwe, T., & Geles, W. (2020) · Genetic and Evolutionary Computation Conference Companion (GECCO Companion) · doi:10.1145/3377929.3389950 · Cited by 3
“probabilistic adaptation technique to provide fault recovery in legged robots [5].”
Reinforcement Learning Compensator Robust to the Time Constants of First Order Delay Elements
Kobayashi, S., & Shibuya, T. (2020) · Conference proceedings/Conference proceedings - IEEE International Conference on Systems, Man, and Cybernetics · doi:10.1109/smc42975.2020.9283188 · Cited by 1
“Such reinforcement learning is intended to reduce the burden on designers by automatically securing the robot’s control rules [1].”
BackgroundA deep dive into exploring the preference hypervolume
Hagg, A., Asteroth, A., & Bäck, T. (2020) · Publication Server of Bonn-Rhein-Sieg University of Applied Sciences (Bonn-Rhein-Sieg University of Applied Sciences) · Cited by 3
Artificial Emotions in Intelligent machines
Shiller, A. V. (2021) · Artificial Societies · doi:10.18254/s207751800014145-8 · Cited by 1
Automated exploration of DNA-based structure self-assembly networks
Cazenille, L., Baccouche, A., & Aubert-Kato, N. (2021) · Royal Society Open Science · doi:10.1098/rsos.210848 · Cited by 1
“These algorithms were previously used to efficiently explore the space of a variety of problems in the robotics [30,31], reinforcement learning [32 – 34] and video games [35,36] communities, but they were also used to explore problems in biology and chemistry [25,37,38].”
MethodSafe Locomotion Within Confined Workspace using Deep Reinforcement Learning
Dastider, A., Raza, S. J. A., & Lin, M. (2021) · IEEE International Conference on Robotic Computing (IRC) · doi:10.1109/irc52146.2021.00025 · Cited by 1
“For developing versatile locomotion technique in animal-like robots, in [8] repertoire-inspired pre-obtained gait library has been utilized to guide robotic locomotion and action planning.”
MethodFCP Gait Control for Hexapod Robot Capable of Decreasing/Increasing Number of Walking Legs
Hosogaya, H., Inagaki, S., & Suzuki, T. (2022) · Transactions of the Society of Instrument and Control Engineers · doi:10.9746/sicetr.58.304 · Cited by 1
Machine Learning for Software-Defined Networking
Guo, Z. (2022) · SpringerBriefs in computer science · doi:10.1007/978-981-19-4874-9_1 · Cited by 1
Strappa, J., Caymes‐Scutari, P., & Bianchini, G. (2022) · Algorithms · doi:10.3390/a15120478 · Cited by 1
Pontin, M., & Damian, D. D. (2023) · IEEE Robotics and Automation Letters · doi:10.1109/lra.2023.3243440 · Cited by 1
Effective Skill Learning via Autonomous Goal Representation Learning
Tsakonas, C. G., & Chatzilygeroudis, K. I. (2023) · International Conference on Information, Intelligence, Systems & Applications (IISA) · doi:10.1109/iisa59645.2023.10345879 · Cited by 1
“For this task, we turn to QD algorithms as they have been successfully used to find both diverse and high-performing behaviors even for complex continuous tasks [16], [17], [20]–[22].”
BackgroundMethodBayesian Optimization for the Inverse Problem in Electrocardiography
Lopez‐Rincon, A., Rojas-Velázquez, D., Garssen, J., van der Laan, S. W., Oberski, D. L., & Tonda, A. P. (2023) · IEEE Symposium Series on Computational Intelligence (SSCI) · doi:10.1109/ssci52147.2023.10371791 · Cited by 1
“Such algorithms are well-suited for problems where using only few evaluations is critical, for example robotics and machine learning [33]–[36].”
BackgroundA Self-Organization Strategy for Unmanned Autonomous Systems
Lee, B., Oh, S., Balchanos, M. G., & Vachtsevanos, G. (2017) · PHM Society Asia-Pacific Conference · doi:10.36001/phmap.2017.v1i1.1808 · Cited by 1
“Cully et al. (2015) suggest an improved trial and error method to determine the optimal action for a walking hexapod with a broken leg, but the large original search space and minutes of lengthy adaptation time to the next step hamper their development.”
MethodCovariance matrix adaptation MAP-Elites for video game level generation
Sun, M., Jiang, G., & Zhang, Y. (2024) · International Conference on Computer Graphics, Artificial Intelligence, and Data Processing (ICCAID) · doi:10.1117/12.3026544 · Cited by 1
Robustness in Diversity: Quality-Diversity Driven Discovery of Agile Soft Robotic Gaits
Daly, J., Casper, D., Farooq, M. U., James, A., Khan, A. N., Mulgrew, P., Tyebkhan, D., Vo, B., & Rieffel, J. A. (2024) · IEEE International Conference on Soft Robotics (RoboSoft) · doi:10.1109/robosoft60065.2024.10521943 · Cited by 1
“We use it in our work as a baseline optimizer against which to compare our QDA approach. b) Quality Diversity Algorithms: Unlike single-objective optimizers like CMA-ES, Quality Diversity algorithms (QDAs) [15–18] search for a combination of both novelty and quality .”
BackgroundMethodLabartkava, G., Suhaimi, N. B., Silva, T. H., & Kirischian, L. (2024) · IEEE World AI IoT Congress (AIIoT) · doi:10.1109/aiiot61789.2024.10579027 · Cited by 1
“There are various robots that currently exist, capable of autonomously or manually changing shape and completely reconfiguring [8-14].”
BackgroundMethodResultA Quality Diversity Study in EvoDevo Processes for Engineering Design
Buchanan, E., Hickinbotham, S., Dubey, R., Friel, I., Colligan, A. R., Price, M. A., & Tyrrell, A. M. (2024) · IEEE Congress on Evolutionary Computation (CEC) · doi:10.1109/cec60901.2024.10612076 · Cited by 1
“The motivation for using QD within an EvoDevo algorithm for design is similar to that for using QD in evolutionary robotics applications [6]–[8]: the space of possible designs is very large and highly heterogeneous, so driving diversity in the population is a potential means to overcome local…”
MethodTowards a Unified Framework for Software-Hardware Integration in Evolutionary Robotics
Buchanan, E., Goff, L. K. L., Hale, M. F., Hart, E., Eiben, A. E., Carlo, M. D., Angus, M., Woolley, R., Timmis, J., Winfield, A. F. T., et al. (2024) · Robotics · doi:10.3390/robotics13110157 · Cited by 1
“This is because digital evolution offers the advantage of rapidly evaluating a wide range of diverse robot designs and controllers without the risk of damaging physical components, and it avoids wasting resources on fabricating suboptimal robots [3,11,12].”
InfluentialDing, B., Kırtay, M., & Spigler, G. (2024) · IEEE-RAS International Conference on Humanoid Robots (Humanoids) · doi:10.1109/humanoids58906.2024.10769814 · Cited by 1
“We took an approach similar to previous work on motion-capture trajectory modeling [2]–[4], and chose to model human head-movement trajectories using a variational au-toencoder (VAE) [18].”
BackgroundMethodSIMULATION OF THE MOVEMENT OF A WALKING ROBOT WITH A DAMAGED LIMB IN MATLAB
Korolenko, Y. R., & Pavlovska, Y. O. (2024) · Scientific notes of Taurida National V I Vernadsky University Series Technical Sciences · doi:10.32782/2663-5941/2024.5.1/25 · Cited by 1
Walking Small Unmanned Vehicles
Pavlovskyi, O. (2025) · Studies in systems, decision and control · doi:10.1007/978-3-031-82035-9_11 · Cited by 1
Scaling Policy Gradient Quality-Diversity with Massive Parallelization via Behavioral Variations
Mitsides, K., Faldor, M., & Cully, A. (2025) · Genetic and Evolutionary Computation Conference (GECCO) · doi:10.1145/3712256.3726335 · Cited by 1
“Nevertheless, its reliance on genetic variation operators lead to slow convergence in expansive search spaces, making it less e ! ective for evolving large deep neural networks (DNNs) [7, 8, 30].”
Self-citationLeung, B., Haomachai, W., Pedersen, J. W., Risi, S., & Manoonpong, P. (2025) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros60139.2025.11247433 · Cited by 1
“Cully et al. [33] used quality diversity algorithms to learn a repertoire of different gaits.”
Collective Dynamics in Physical and Social Networks
Isakov, A. (2016) · Digital Access to Scholarship at Harvard (DASH) (Harvard University) · Cited by 1
Collective behaviour of zibrafish and robot groups in a constrained environment.
Chemtob, Y. (2020) · theses.fr (ABES) · doi:10.70675/53828f4dza8b9z404bza417zb8903b142b78 · Cited by 1
Divergent Search for Few-Shot Image ClassificationPreprint
Tan, J., & Kainz, B. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2004.07903 · Cited by 1
“This endows the robot with a range of pre-learned skills making it better prepared for uncertain terrain or new, unseen tasks [6].”
BackgroundRobot Design With Neural Networks, MILP Solvers and Active LearningPreprint
Narain, S., Mak, E., Chee, D., Huster, T. P., Cohen, J., POCHIRAJU, K. V., Englot, B., Jha, N. K., & Narayan, K. S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2010.09842 · Cited by 1
“Constraints are represented naturally as constraints in an MILP language, not indirectly as hard-to-design penalty or barrier functions [6, 5] as with constrained optimization methods [6,7,8].”
BackgroundMethodInfluentialOn the use of feature-maps and parameter control for improved quality-diversity meta-evolutionPreprint
Bossens, D. M., & Tarapore, D. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2105.10317 · Cited by 1
“Prior work has shown that performing a search across the behaviour-performance maps evolved by MAP-Elites can give high-performing recovery solutions within a limited number of function evaluations (Cully et al., 2015).”
BackgroundSelf-citationFrans, K., Soros, L. B., & Witkowski, O. (2021) · arXiv (Cornell University) · doi:10.48550/arxiv.2106.09153 · Cited by 1
“Since then, a variety of algorithms have been developed to combine selection for both fitness and diversity, including novelty search with local competition (Lehman and Stanley, 2011b) and MAP-Elites (Cully et al., 2015).”
BackgroundLearning in Sparse Rewards settings through Quality-Diversity algorithmsPreprint
Paolo, G. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2203.01027 · Cited by 3
A proposal to improve research in AI algorithm and data governancePreprint
Braga, J., Regateiro, F. S., Stiubiener, I., & Braga, J. C. (2022) · OSF Preprints (OSF Preprints) · doi:10.31219/osf.io/sr7kt · Cited by 1
Improving Robotic Decision-Making in Unmodeled Situations
Ohi, N. (2022) · The Research Repository @ WVU (West Virginia University) · doi:10.33915/etd.11537 · Cited by 1
AI-driven Automated Discovery Tools Reveal Diverse Behavioral Competencies of Biological NetworksPreprint
Etcheverry, M., Moulin-Frier, C., Oudeyer, P., & Levin, M. G. (2024) · eLife · doi:10.7554/elife.92683.1 · Cited by 1
Basil: Best-Action Symbolic Interpretable Learning for Evolving Compact Rl PoliciesPreprint
Shahnazari, K., Ayyoubzadeh, S. M., & Keshtparvar, M. (2025) · SSRN Electronic Journal · doi:10.2139/ssrn.5364978 · Cited by 1
Buzz, Choose, Forget: A Meta-Bandit Framework for Bee-Like Decision MakingPreprint
Claeys, E., Kerjean, E., & Loubes, J. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2510.16462 · Cited by 1
Joint-Local Grounded Action Transformation for Sim-to-Real Transfer in Multi-Agent Traffic ControlPreprint
Turnau, J., Da, L., Vo, K., Rafi, F. A., Bachiraju, S., Chen, T., & Wei, H. (2025) · arXiv (Cornell University) · doi:10.48550/arxiv.2507.15174 · Cited by 5
Pontes, J. P. D. C. (2019) · theses.fr (ABES) · doi:10.70675/dd3a2f35z6ac7z465cz8735za40cfd594da1 · Cited by 1
IDEAgent: Agentic Quality-Diversity Search for Research Idea GenerationPreprint
Gumma, V., Majumder, N., Sinhahajari, S., & Poria, S. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2607.22375 · Cited by 3
“Standard QD methods organize the search space using predefined axes and cells (Bradley et al., 2024; Lehman et al., 2024; Lehman and Stanley, 2011; Mouret and Clune, 2015; Pugh et al., 2016; Fontaine and Nikolaidis, 2021; Cully et al., 2015).”
Toward Recursive Self-Improvement in Physical AIPreprint
Luo, M., Bai, Z., Xu, S., Li, Y., Li, K., Qian, M., Lu, W., Zhang, R., Zou, Y., Luo, Z., et al. (2026) · arXiv (Cornell University) · doi:10.48550/arxiv.2610.29677 · Cited by 2
Toward Enactive Artificial Intelligence
Rafiee, B., & Sutton, R. (2026) · arXiv · doi:10.48550/arxiv.2605.24238
“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).”
Zhang, J., Zhao, B., Yang, W., Foerster, J., Clune, J., Jiang, M., Devlin, S., & Shavrina, T. (2026) · arXiv · doi:10.48550/arxiv.2603.19461
Self-citationHeuresis: Search Strategies for Autonomous AI Research Agents Across Quality, Diversity and Novelty
Antoniades, A., Nathani, D., Saha, R., Amayuelas, A., Bercovich, I., Weng, Z., Baskaran, V., Bhatia, K., & Wang, W. Y. (2026) · arXiv · doi:10.48550/arxiv.2606.25198
The man-machine integration eraPreprint
Bilar, D. (2016) · doi:10.7287/peerj.preprints.2402v1
“Skills Projects Experimental Design CMU 15-321: Research Methods for Experimental Computer Science [45] Information Theory MICMINE: Maximal information-based nonparametric exploration statistics [46] Control and Optimization Remy: Computer-generated, adaptive TCP congestion control algorithms [47] Modeling and Invaria…”
MethodLose a Leg but not Your Head – A Cognitive Extension of a Biologically-inspired Walking Architecture
Schilling, M. (2016) · Procedia Computer Science · doi:10.1016/j.procs.2016.07.412
Chairs' welcome for GECCO'17 workshop "evolution in cognition"
Doncieux, S., Auerbach, J. E., Duro, R. J., & de Vladar, H. P. (2017) · Genetic and Evolutionary Computation Conference Companion (GECCO Companion) · doi:10.1145/3067695.3082462
“In our group, we work on learning algorithms that allow robots to adapt more like animals, that is, in a few minutes [3].”
BackgroundEvolution of Decision-Making Systems
Schossau, J. D. (2017) · Michigan State University Libraries · doi:10.25335/m5wf78
A baseline-realistic objective open-ended kinematics simulator for evolutionary robotics
Konsella, R., Chiarulli, F., Peterson, J. R., & Rieffel, J. A. (2017) · Genetic and Evolutionary Computation Conference Companion (GECCO Companion) · doi:10.1145/3067695.3082164 · Cited by 1
“…exist to trying to "cross" the reality gap - beginning with Jakobi’s own modeling of sensor noise [12], and including more sophisticated techniques such as Bongard’s co-evolutionary self-modeling [1], Cully et al. ’s Bayesian-Optimization approach [6], and Mouret’s transferability approach [13].”
MethodDas ICH, die Optimierung und die Gesellschaft
Küppers, E. W. U. (2017) · Springer eBooks · doi:10.1007/978-3-658-17920-5_4
An insect-computer hybrid walking robot
Cao, F. (2018) · DR-NTU (Nanyang Technological University) · doi:10.32657/10356/74213
Expressivity in Natural and Artificial SystemsPreprint
LaViers, A. (2018) · arXiv (Cornell University) · doi:10.48550/arxiv.1807.02016
Rapid and thorough exploration of low dimensional phenotypic landscapes
Smith, D. L. (2017) · Queen Mary Research Online (Queen Mary University of London)
Autonomous robots learning and social integration on the basis of modern cognitive technologies
Diane, S. A. K. (2018) · Philosophy of Science and Technology · doi:10.21146/2413-9084-2018-23-2-89-102
'Instinctive' robot recovers from injury fast
Castelvecchi, D. (2015) · Nature · doi:10.1038/nature.2015.17641
Hierarchical Surrogate Modeling for Illumination AlgorithmsPreprint
Hagg, A. (2017) · arXiv (Cornell University) · doi:10.48550/arxiv.1703.09926
“Today, optimization of high-dimensional problems, such as in computational fluid dynamics [8] or robotics [7], can be extremely expensive in terms of computational effort.”
BackgroundMethodScalable Co-Optimization Of Morphology And Control In Embodied Machines
Cheney, N., Bongard, J., SunSpiral, V., & Lipson, H. (2017) · Zenodo (CERN European Organization for Nuclear Research) · doi:10.5281/zenodo.267998
Damage Recovery for Simulated Modular Robots Through Joint Evolution of Morphologies and Controllers
Akrour, D., & Djedi, N. (2019) · Journal of Automation Mobile Robotics & Intelligent Systems · doi:10.14313/jamris_1-2019/2 · Cited by 2
“Building resilient robots capable of recovering from damages is a central and challenging question [1, 2].”
BackgroundEPANer Team Description Paper for World Robot Challenge 2020Preprint
Yan, Z., Crombez, N., & Sun, L. (2019) · arXiv (Cornell University) · doi:10.48550/arxiv.1909.02355
“For the last aspect, we intend to adopt an approach similar to the behavior-performance map [8].”
Impact des interventions assistées d’un animal sur le résident
Chrusciel, J., Letty, A., Armaingaud, D., Haÿ, P.-É., & Sanchez, S. (2019) · Soins Gérontologie · doi:10.1016/j.sger.2019.09.003
Training Neural Networks Through the Integration of Evolution and Gradient Descent
Morse, G. (2019) · Journal of International Crisis and Risk Communication Research · Cited by 2
Jutharee, W., Kaewkamnerdpong, B., Polvichai, J., & Maneewarn, T. (2019) · International Conference on Control, Automation and Systems (ICCAS) · doi:10.23919/iccas47443.2019.8971707
Self Modeling and Gait Control of Quadruped Robot Using Q-Learning Based Particle Swarm Optimization
Meerza, S. I. A., & Uzzal, M. M. (2019) · AIUB Journal of Science and Engineering (AJSE) · doi:10.53799/ajse.v18i3.24
Deep learning methods for style extraction and transfer
Mohammed, O. (2019) · theses.fr (ABES) · doi:10.70675/6197d769z096cz4d9fz90c2zf30876ff6a39
Natural Language Interaction to Facilitate Mental Models of Remote Robots
Garcia, F. J. C., Lopes, J., & Hastie, H. (2020) · Edinburgh Research Explorer (University of Edinburgh) · doi:10.48550/arxiv.2003.05870 · Cited by 1
“This problem is confounded by the advent of robots that can adapt, with their functionality and behaviour changing continuously [5].”
Open-ended Search through Minimal Criterion Coevolution
Brant, J. C. (2020) · STARS (University of Central Florida)
Policy Manifold Search for Improving Diversity-based NeuroevolutionPreprint
Rakicevic, N., Cully, A., & Kormushev, P. S. (2020) · arXiv (Cornell University) · doi:10.48550/arxiv.2012.08676
“[12] Nemanja Rakicevic and Petar Kormushev.”
BackgroundInfluentialSelf-citationDeep learning that scales: leveraging compute and data
Camúñez, V. C. (2020) · UPCommons institutional repository (Universitat Politècnica de Catalunya) · doi:10.5821/dissertation-2117-335426
Fault Diagnosis Method of Brushless Motor Inverter for Substation Inspection Robot
Zhang, X., Lv, Q., Dang, X., & Liu, S. (2020) · International Forum on Electrical Engineering and Automation (IFEEA) · doi:10.1109/ifeea51475.2020.00062
“Breakdown faults are generally protected by hardware circuits and converted into open circuit faults[5-6].”
BackgroundOn the use of feature-maps for improved quality-diversity meta-evolution
Bossens, D. M., & Tarapore, D. (2021) · Genetic and Evolutionary Computation Conference Companion (GECCO Companion) · doi:10.1145/3449726.3459442
“Prior work has shown that performing a search across the behaviour-performance maps evolved by MAP-Elites can give high-performing recovery solutions within a limited number of function evaluations (Cully et al., 2015).”
BackgroundSelf-citationEskandari, F., Bahramizadeh-Sajadi, S., & Katoozian, H. R. (2021) · Journal of Orthopedic and Spine Trauma · doi:10.18502/jost.v7i2.6999
Model Based Quality Diversity Optimization
Schneider, L. (2021) · Open access LMU (Ludwid Maxmilian's Universitat Munchen) · doi:10.5282/ubm/epub.77443
Coupling-dependent convergence behavior of phase oscillators with tegotae-control
Hauser, S., Dujany, M., Arreguit, J., Ijspeert, A. J., & Iida, F. (2021) · IEEE/RSJ International Conference on Intelligent Robots and Systems (IROS) · doi:10.1109/iros51168.2021.9636646
“They range from building a gait database in simulation to include possible hardware modifications [1], reinforcement learning techniques for sim-to-real transfer of controllers [2], synthesizing models from sensory information in the real world [3], and model-free evolutionary algorithms co-evolving the morphology and…”
BackgroundBotha, B., & du Plessis, M. C. (2022) · Communications in computer and information science · doi:10.1007/978-3-030-95070-5_15
Artificial Intelligence model for social sustainability risk management in the apparel supply chainPreprint
Bubicz, M. E. (2022) · doi:10.31224/2209
Diagnosing Deep SLAM for Domain-Shift Localization
Yoshida, M., & Kanji, T. (2022) · IEEE/SICE International Symposium on System Integration (SII) · doi:10.1109/sii52469.2022.9708872 · Cited by 1
Knowledge-Integrated Informed AI for National SecurityPreprint
Myne, A., Leahy, K., & Soklaski, R. (2022) · arXiv (Cornell University) · doi:10.48550/arxiv.2202.03188 · Cited by 1
“For example, in “Robots that can adapt like animals” [39], the authors build a low-dimensional behavioral map of robot performance in simulation during the initial training phase (Figure 22, top row).”
BackgroundFree Simulation Software and Library
Uğurlu, B., & Ivaldi, S. (2018) · Springer eBooks · doi:10.1007/978-94-007-6046-2_27
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.