{"about":{"site":"https://codewithpapers.app","non_affiliation":"Code with Papers and Syntology are not affiliated with, endorsed by, or sponsored by Papers with Code, Meta, or the pwc-archive mirror.","licence":"CC BY-SA 4.0","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","attribution":"https://codewithpapers.app/attribution","modified":"archive material modified by Syntology; see the attribution page"},"url":"/paper/setting-up-a-reinforcement-learning-task-with","title":"Setting up a Reinforcement Learning Task with a Real-World Robot","arxiv_id":"1803.07067","date":"2018-03-19","proceeding":null,"authors":["A. Rupam Mahmood","Dmytro Korenkevych","Brent J. Komer","James Bergstra"],"abstract":"Reinforcement learning is a promising approach to developing hard-to-engineer\nadaptive solutions for complex and diverse robotic tasks. However, learning\nwith real-world robots is often unreliable and difficult, which resulted in\ntheir low adoption in reinforcement learning research. This difficulty is\nworsened by the lack of guidelines for setting up learning tasks with robots.\nIn this work, we develop a learning task with a UR5 robotic arm to bring to\nlight some key elements of a task setup and study their contributions to the\nchallenges with robots. We find that learning performance can be highly\nsensitive to the setup, and thus oversights and omissions in setup details can\nmake effective learning, reproducibility, and fair comparison hard. Our study\nsuggests some mitigating steps to help future experimenters avoid difficulties\nand pitfalls. We show that highly reliable and repeatable experiments can be\nperformed in our setup, indicating the possibility of reinforcement learning\nresearch extensively based on real-world robots.","url_abs":"http://arxiv.org/abs/1803.07067v1","url_pdf":"http://arxiv.org/pdf/1803.07067v1.pdf","source":{"archive":"pwc-archive (Hugging Face), CC BY-SA 4.0","snapshot":"2025-07-28","licence_url":"https://creativecommons.org/licenses/by-sa/4.0/legalcode","row_kind":"abstracts"},"code_links":[{"paper_slug":"setting-up-a-reinforcement-learning-task-with","repo_url":"https://github.com/dti-research/SenseAct","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"BSD-3-Clause"}},{"paper_slug":"setting-up-a-reinforcement-learning-task-with","repo_url":"https://github.com/kindredresearch/SenseAct","is_official":0,"mentioned_in_paper":0,"mentioned_in_github":1,"framework":"none","reach":{"status":"ok","spdx":"BSD-3-Clause"}}],"tasks":[{"task_slug":"reinforcement-learning","task_name":"Reinforcement Learning"},{"task_slug":"reinforcement-learning-1","task_name":"Reinforcement Learning (RL)"},{"task_slug":"reinforcement-learning-2","task_name":"reinforcement-learning"}],"methods":[],"datasets_introduced":[],"methods_introduced":[],"results":[],"syntology":{"atlas_url":"https://app.syntology.ai/?focus=1803.07067","mcp":null,"developers":"https://syntology.ai/developers"},"arxiv_metadata":null,"syntology_extracted_results":null}