{"status":"ok","message-type":"work","message-version":"1.0.0","message":{"indexed":{"date-parts":[[2025,12,23]],"date-time":"2025-12-23T10:04:23Z","timestamp":1766484263224,"version":"3.37.3"},"reference-count":37,"publisher":"MIT Press","issue":"2","content-domain":{"domain":["direct.mit.edu"],"crossmark-restriction":true},"short-container-title":[],"published-print":{"date-parts":[[2024,6,3]]},"abstract":"<jats:title>Abstract<\/jats:title><jats:p>In this paper, we compare Bayesian Optimization, Differential Evolution, and an Evolution Strategy employed as a gait-learning algorithm in modular robots. The motivational scenario is the joint evolution of morphologies and controllers, where \u201cnewborn\u201d robots also undergo a learning process to optimize their inherited controllers (without changing their bodies). This context raises the question: How do gait-learning algorithms compare when applied to various morphologies that are not known in advance (and thus need to be treated as without priors)? To answer this question, we use a test suite of twenty different robot morphologies to evaluate our gait-learners and compare their efficiency, efficacy, and sensitivity to morphological differences. The results indicate that Bayesian Optimization and Differential Evolution deliver the same solution quality (walking speed for the robot) with fewer evaluations than the Evolution Strategy. Furthermore, the Evolution Strategy is more sensitive for morphological differences (its efficacy varies more between different morphologies) and is more subject to luck (repeated runs on the same morphology show greater variance in the outcomes).<\/jats:p>","DOI":"10.1162\/evco_a_00334","type":"journal-article","created":{"date-parts":[[2023,5,18]],"date-time":"2023-05-18T16:49:51Z","timestamp":1684428591000},"page":"105-124","update-policy":"https:\/\/doi.org\/10.1162\/mitpressjournals.corrections.policy","source":"Crossref","is-referenced-by-count":3,"title":["Comparing Robot Controller Optimization Methods on Evolvable Morphologies"],"prefix":"10.1162","volume":"32","author":[{"ORCID":"https:\/\/orcid.org\/0000-0002-7972-1649","authenticated-orcid":true,"given":"Fuda","family":"van Diggelen","sequence":"first","affiliation":[{"name":"Department of Computer Science, Vrije Universiteit Amsterdam, The Netherlands fuda.van.diggelen@vu.nl"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-2213-8356","authenticated-orcid":true,"given":"Eliseo","family":"Ferrante","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Vrije Universiteit Amsterdam, The Netherlands Technology Innovation Institute, Abu Dhabi, P.O. Box: 9639, Masdar City, UAE e.ferrante@vu.nl"}],"role":[{"role":"author","vocabulary":"crossref"}]},{"ORCID":"https:\/\/orcid.org\/0000-0002-3106-4213","authenticated-orcid":true,"given":"A. E.","family":"Eiben","sequence":"additional","affiliation":[{"name":"Department of Computer Science, Vrije Universiteit Amsterdam, The Netherlands a.e.eiben@vu.nl"}],"role":[{"role":"author","vocabulary":"crossref"}]}],"member":"281","published-online":{"date-parts":[[2024,6,3]]},"reference":[{"volume-title":"Simulated annealing and Boltzmann machines: A stochastic approach to combinatorial optimization and neural computing","year":"1989","author":"Aarts","key":"2024060313544264700_B1"},{"key":"2024060313544264700_B2","first-page":"136","article-title":"Robogen: Robot generation through artificial evolution","author":"Auerbach","year":"2014","journal-title":"Artificial Life Conference Proceedings 14"},{"issue":"8","key":"2024060313544264700_B3","doi-asserted-by":"publisher","first-page":"74","DOI":"10.1145\/2493883","article-title":"Evolutionary robotics","volume":"56","author":"Bongard","year":"2013","journal-title":"Communications of the ACM"},{"key":"2024060313544264700_B4","first-page":"226","article-title":"On the difficulty of co-optimizing morphology and control in evolved virtual creatures","author":"Cheney","year":"2016","journal-title":"Artificial Life Conference Proceedings 13"},{"issue":"143","key":"2024060313544264700_B5","doi-asserted-by":"publisher","DOI":"10.1098\/rsif.2017.0937","article-title":"Scalable co-optimization of morphology and control in embodied machines","volume":"15","author":"Cheney","year":"2018","journal-title":"Journal of the Royal Society Interface"},{"key":"2024060313544264700_B6","first-page":"2272","article-title":"Influences of artificial speciation on morphological robot evolution","author":"De Carlo","year":"2020","journal-title":"IEEE Symposium Series on Computational Intelligence"},{"issue":"4","key":"2024060313544264700_B7","doi-asserted-by":"publisher","DOI":"10.3389\/frobt.2015.00004","article-title":"Evolutionary robotics: What, why, and where to","volume":"2","author":"Doncieux","year":"2015","journal-title":"Frontiers in Robotics and AI"},{"key":"2024060313544264700_B8","first-page":"1056","article-title":"The triangle of life: Evolving robots in real-time and real-space","author":"Eiben","year":"2013","journal-title":"Artificial Life Conference Proceedings 13"},{"key":"2024060313544264700_B9","first-page":"1383","article-title":"If it evolves it needs to learn","author":"Eiben","year":"2020","journal-title":"Proceedings of the Genetic and Evolutionary Computation Conference Companion (GECCO)"},{"key":"2024060313544264700_B10","doi-asserted-by":"crossref","DOI":"10.1007\/978-3-662-05094-1","volume-title":"Introduction to evolutionary computing","author":"Eiben","year":"2003"},{"key":"2024060313544264700_B11","doi-asserted-by":"crossref","first-page":"1423","DOI":"10.1007\/978-3-540-30301-5_62","article-title":"Evolutionary Robotics","volume-title":"Handbook of robotics","author":"Floreano","year":"2008"},{"key":"2024060313544264700_B12","doi-asserted-by":"crossref","first-page":"997","DOI":"10.1145\/1276958.1277158","article-title":"Generating large-scale neural networks through discovering geometric regularities","author":"Gauci","year":"2007","journal-title":"Proceedings of the Genetic and Evolutionary Computation Conference (GECCO)"},{"article-title":"The CMA evolution strategy: A tutorial","year":"2016","author":"Hansen","key":"2024060313544264700_B13"},{"key":"2024060313544264700_B14","first-page":"868","article-title":"Body-brain co-evolution using L-systems as a generative encoding","author":"Hornby","year":"2001","journal-title":"Proceedings of the 3rd Annual Conference on Genetic and Evolutionary Computation"},{"issue":"4","key":"2024060313544264700_B15","doi-asserted-by":"publisher","first-page":"642","DOI":"10.1016\/j.neunet.2008.03.014","article-title":"Central pattern generators for locomotion control in animals and robots: A review","volume":"21","author":"Ijspeert","year":"2008","journal-title":"Neural Networks"},{"key":"2024060313544264700_B16","first-page":"849","article-title":"Policy search for motor primitives in robotics","author":"Kober","year":"2009","journal-title":"Advances in Neural Information Processing Systems"},{"article-title":"Learning directed locomotion in modular robots with evolvable morphologies","year":"2020","author":"Lan","key":"2024060313544264700_B17"},{"article-title":"Time efficiency in optimization with a Bayesian-Evolutionary Algorithm","year":"2020","author":"Lan","key":"2024060313544264700_B18"},{"key":"2024060313544264700_B19","first-page":"432","article-title":"Sample and time efficient policy learning with CMA-ES and Bayesian Optimisation","author":"Le Goff","year":"2020","journal-title":"Artificial Life Conference Proceedings"},{"key":"2024060313544264700_B20","doi-asserted-by":"publisher","DOI":"10.1109\/TCDS.2022.3148543","article-title":"Morpho evolution with learning using a controller archive as an inheritance mechanism","author":"Le Goff","year":"2022","journal-title":"IEEE Transactions on Cognitive and Developmental Systems"},{"key":"2024060313544264700_B21","doi-asserted-by":"crossref","DOI":"10.3389\/frobt.2022.797393","article-title":"The effects of learning in morphologically evolving robot systems","volume":"9","author":"Luo","year":"2022","journal-title":"Frontiers in Robotics and AI"},{"key":"2024060313544264700_B22","doi-asserted-by":"crossref","first-page":"129","DOI":"10.1145\/3449639.3459315","article-title":"Biodiversity in evolved voxel-based soft robots","author":"Medvet","year":"2021","journal-title":"Proceedings of the Genetic and Evolutionary Computation Conference"},{"key":"2024060313544264700_B23","doi-asserted-by":"publisher","first-page":"86","DOI":"10.1007\/978-3-030-43722-0_6","article-title":"Evolving-controllers versus learning-controllers for morphologically evolvable robots","volume":"12104","author":"Miras","year":"2020","journal-title":"Applications of Evolutionary Computation"},{"key":"2024060313544264700_B24","first-page":"224","article-title":"Effects of selection preferences on evolved robot morphologies and behaviors","author":"Miras","year":"2018","journal-title":"Artificial Life Conference Proceedings"},{"key":"2024060313544264700_B25","doi-asserted-by":"crossref","first-page":"703","DOI":"10.1007\/978-3-319-77538-8_47","article-title":"Search space analysis of evolvable robot morphologies","author":"Miras","year":"2018","journal-title":"International Conference on the Applications of Evolutionary Computation"},{"key":"2024060313544264700_B26","doi-asserted-by":"crossref","DOI":"10.7551\/mitpress\/2889.001.0001","volume-title":"Evolutionary robotics: The biology, intelligence, and technology of self-organizing machines","author":"Nolfi","year":"2000"},{"key":"2024060313544264700_B27","doi-asserted-by":"crossref","first-page":"825","DOI":"10.1007\/978-3-319-55849-3_53","article-title":"Overcoming initial convergence in multi-objective evolution of robot control and morphology using a two-phase approach","author":"Nygaard","year":"2017","journal-title":"European Conference on the Applications of Evolutionary Computation"},{"key":"2024060313544264700_B28","doi-asserted-by":"crossref","first-page":"37","DOI":"10.1109\/EH.2000.869340","article-title":"The golem project: Evolving hardware bodies and brains","author":"Pollack","year":"2000","journal-title":"Proceedings of the Second NASA\/DoD Workshop on Evolvable Hardware"},{"issue":"12","key":"2024060313544264700_B29","doi-asserted-by":"crossref","first-page":"1804","DOI":"10.1017\/S0263574718000747","article-title":"A survey on evolutionary-aided design in robotics","volume":"36","author":"Prabhu","year":"2018","journal-title":"Robotica"},{"key":"2024060313544264700_B30","first-page":"2951","article-title":"Practical Bayesian optimization of machine learning algorithms","volume":"2","author":"Snoek","year":"2012","journal-title":"Proceedings of the 25th International Conference on Neural Information Processing Systems"},{"issue":"2","key":"2024060313544264700_B31","doi-asserted-by":"publisher","first-page":"99","DOI":"10.1162\/106365602320169811","article-title":"Evolving neural networks through augmenting topologies","volume":"10","author":"Stanley","year":"2002","journal-title":"Evolutionary Computation"},{"issue":"4","key":"2024060313544264700_B32","doi-asserted-by":"publisher","first-page":"341","DOI":"10.1023\/A:1008202821328","article-title":"Differential evolution\u2014A simple and efficient heuristic for global optimization over continuous spaces","volume":"11","author":"Storn","year":"1997","journal-title":"Journal of Global Optimization"},{"key":"2024060313544264700_B33","doi-asserted-by":"crossref","DOI":"10.1145\/3377929.3389972","article-title":"Differential evolution with reversible linear transformations","author":"Tomczak","year":"2020"},{"key":"2024060313544264700_B34","first-page":"2117","article-title":"The effects of adaptive control on learning directed locomotion","author":"van Diggelen","year":"2020","journal-title":"IEEE Symposium Series on Computational Intelligence"},{"key":"2024060313544264700_B35","doi-asserted-by":"crossref","DOI":"10.7551\/mitpress\/8493.001.0001","volume-title":"The horizons of evolutionary robotics","author":"Vargas","year":"2014"},{"issue":"1","key":"2024060313544264700_B36","doi-asserted-by":"publisher","first-page":"80","DOI":"10.1162\/ARTL_a_00223","article-title":"Online gait learning for modular robots with arbitrary shapes and sizes","volume":"23","author":"Weel","year":"2017","journal-title":"Artificial Life"},{"key":"2024060313544264700_B37","doi-asserted-by":"crossref","first-page":"331","DOI":"10.1007\/978-3-662-44616-4_13","article-title":"Evolvable robot hardware","volume-title":"Evolvable hardware","author":"Winfield","year":"2015"}],"container-title":["Evolutionary Computation"],"original-title":[],"language":"en","link":[{"URL":"https:\/\/direct.mit.edu\/evco\/article-pdf\/32\/2\/105\/2372825\/evco_a_00334.pdf","content-type":"application\/pdf","content-version":"vor","intended-application":"syndication"},{"URL":"https:\/\/direct.mit.edu\/evco\/article-pdf\/32\/2\/105\/2372825\/evco_a_00334.pdf","content-type":"unspecified","content-version":"vor","intended-application":"similarity-checking"}],"deposited":{"date-parts":[[2024,10,20]],"date-time":"2024-10-20T18:28:59Z","timestamp":1729448939000},"score":1,"resource":{"primary":{"URL":"https:\/\/direct.mit.edu\/evco\/article\/32\/2\/105\/116038\/Comparing-Robot-Controller-Optimization-Methods-on"}},"subtitle":[],"short-title":[],"issued":{"date-parts":[[2024]]},"references-count":37,"journal-issue":{"issue":"2","published-online":{"date-parts":[[2024,6,3]]},"published-print":{"date-parts":[[2024,6,3]]}},"URL":"https:\/\/doi.org\/10.1162\/evco_a_00334","relation":{},"ISSN":["1530-9304"],"issn-type":[{"type":"electronic","value":"1530-9304"}],"subject":[],"published-other":{"date-parts":[[2024]]},"published":{"date-parts":[[2024]]}}}