Felix Wiebe, Shivesh Kumar, Lasse Shala, S. Vyas, M. Javadi, Frank Kirchner
{"title":"Open Source Dual-Purpose Acrobot and Pendubot Platform: Benchmarking Control Algorithms for Underactuated Robotics","authors":"Felix Wiebe, Shivesh Kumar, Lasse Shala, S. Vyas, M. Javadi, Frank Kirchner","doi":"10.1109/MRA.2023.3341257","DOIUrl":null,"url":null,"abstract":"Recent interest in the control of underactuated robots has surged significantly due to the impressive athletic behaviors shown by robots developed by, e.g., Boston Dynamics (<uri>https://www.bostondynamics.com</uri>), Agility Robotics (<uri>https://agilityrobotics.com/robots</uri>), and the Massachusetts Institute of Technology <xref ref-type=\"bibr\" rid=\"ref1\">[1]</xref>. This gives rise to the need for canonical robotic hardware setups for studying underactuation and comparing learning and control algorithms for their performance and robustness. Similar to OpenAIGym <xref ref-type=\"bibr\" rid=\"ref2\">[2]</xref> and Stable Baselines <xref ref-type=\"bibr\" rid=\"ref3\">[3]</xref>, which provide simulated benchmarking environments and baselines for reinforcement learning algorithms, there is a need for benchmarking learning and control methods on real canonical hardware setups. To encourage reproducibility in robotics and artificial intelligence research, these hardware setups should be affordable and easy to manufacture with off-the-shelf components, and the accompanying software should be open source. Acrobots and pendubots are classical textbook examples of canonical underactuated systems with strong nonlinear dynamics, and their swing-up and upright balancing is considered a challenging control problem, especially on real hardware.","PeriodicalId":55019,"journal":{"name":"IEEE Robotics & Automation Magazine","volume":null,"pages":null},"PeriodicalIF":5.4000,"publicationDate":"2024-06-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"6","resultStr":null,"platform":"Semanticscholar","paperid":null,"PeriodicalName":"IEEE Robotics & Automation Magazine","FirstCategoryId":"94","ListUrlMain":"https://doi.org/10.1109/MRA.2023.3341257","RegionNum":3,"RegionCategory":"计算机科学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"AUTOMATION & CONTROL SYSTEMS","Score":null,"Total":0}
引用次数: 6
Abstract
Recent interest in the control of underactuated robots has surged significantly due to the impressive athletic behaviors shown by robots developed by, e.g., Boston Dynamics (https://www.bostondynamics.com), Agility Robotics (https://agilityrobotics.com/robots), and the Massachusetts Institute of Technology [1]. This gives rise to the need for canonical robotic hardware setups for studying underactuation and comparing learning and control algorithms for their performance and robustness. Similar to OpenAIGym [2] and Stable Baselines [3], which provide simulated benchmarking environments and baselines for reinforcement learning algorithms, there is a need for benchmarking learning and control methods on real canonical hardware setups. To encourage reproducibility in robotics and artificial intelligence research, these hardware setups should be affordable and easy to manufacture with off-the-shelf components, and the accompanying software should be open source. Acrobots and pendubots are classical textbook examples of canonical underactuated systems with strong nonlinear dynamics, and their swing-up and upright balancing is considered a challenging control problem, especially on real hardware.
期刊介绍:
IEEE Robotics & Automation Magazine is a unique technology publication which is peer-reviewed, readable and substantive. The Magazine is a forum for articles which fall between the academic and theoretical orientation of scholarly journals and vendor sponsored trade publications. IEEE Transactions on Robotics and IEEE Transactions on Automation Science and Engineering publish advances in theory and experiment that underpin the science of robotics and automation. The Magazine complements these publications and seeks to present new scientific results to the practicing engineer through a focus on working systems and emphasizing creative solutions to real-world problems and highlighting implementation details. The Magazine publishes regular technical articles that undergo a peer review process overseen by the Magazine''s associate editors; special issues on important and emerging topics in which all articles are fully reviewed but managed by guest editors; tutorial articles written by leading experts in their field; and regular columns on topics including education, industry news, IEEE RAS news, technical and regional activity and a calendar of events.