{"id":"ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","title":"Whole-Body Dexterity lets a humanoid robot use its torso and arms to reposition objects","argument":"Researchers introduce Whole-Body Dexterity (WBD), a system for manipulating objects by using a humanoid robot’s body and appendages—not just its hands—as contact surfaces. The authors train a controller in simulation on different starting and target grasps, then report that it transferred to a real Unitree G1 without additional real-world training. In their reported tests, it repositioned a basketball and box around the torso and completed a task carrying a box from a shelf to a table and placing it. A human teleoperator supplied coarse movement guidance. Across the real-world tasks they tested, the authors report 90% trial completion for WBD, versus 18% for direct retargeting and 4% for a general-purpose whole-body controller. These results suggest a route to handling bulky objects that are awkward to grip with hands alone, but the abstract does not establish how well the approach generalizes beyond the tested robot and tasks.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The evidence is the paper's abstract; it does not provide trial counts, task-by-task results, or evaluation details. The abstract describes one robot model and selected objects and tasks; it does not establish generalization to other robots or settings. The reported controller relies on coarse movement references from a human teleoperator; the abstract does not show fully autonomous task execution.","speaker":{"name":"arXiv eess.SY","roleAtSource":"Publisher"},"source":{"title":"Whole-Body Dexterity","publisher":"arXiv eess.SY","format":"Article","url":"https://arxiv.org/abs/2610.22184","publishedAt":"2026-10-15","dateBasis":"Date basis not provided."},"evidence":[{"text":"The paper presents a framework in which a humanoid robot uses its body and appendages as contact surfaces to grasp and manipulate objects.","excerpt":"uses the robot's body and appendages as contact surfaces to grasp and manipulate objects.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The authors describe training in simulation and report transfer to a real Unitree G1 without additional real-world training.","excerpt":"We show that the distilled policy transfers zero-shot to a real Unitree G1","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The reported real-robot demonstrations include repositioning a basketball and a box, plus carrying a box from a shelf to a table and placing it.","excerpt":"repositions a basketball and a cardboard box to regions around the torso and completes a long-horizon task that stows a box under one arm, carries it from a shelf to a table, and places it.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The controller used coarse motion references from a human teleoperator, so the demonstrations were not described as fully autonomous.","excerpt":"The controller operates interactively, steered by coarse motion references from a human teleoperator.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"Across the reported real-world tasks, the authors say the distilled controller completed 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","excerpt":"Across real-world whole-body manipulation tasks, the distilled controller completes 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"Using the torso and arms as contact surfaces may help with bulky objects that are awkward to handle with hands alone; how broadly this benefit applies is not established by the abstract.","excerpt":"falling short of humans' ability to engage the entire body by bracing and sliding bulky objects against the arms and torso.","url":"https://arxiv.org/abs/2610.22184","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2610.22184","checkedAt":"2026-10-18T21:01:05.022Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"whatWouldChangeView":{"attribution":"Open question","text":"How well would the controller perform on different objects, robots, and tasks without additional training or human guidance?"},"editorialComparisons":[],"live":{"title":"Whole-Body Dexterity lets a humanoid robot use its torso and arms to reposition objects","summary":"Researchers introduce Whole-Body Dexterity (WBD), a system for manipulating objects by using a humanoid robot’s body and appendages—not just its hands—as contact surfaces. The authors train a controller in simulation on different starting and target grasps, then report that it transferred to a real Unitree G1 without additional real-world training. In their reported tests, it repositioned a basketball and box around the torso and completed a task carrying a box from a shelf to a table and placing it. A human teleoperator supplied coarse movement guidance. Across the real-world tasks they tested, the authors report 90% trial completion for WBD, versus 18% for direct retargeting and 4% for a general-purpose whole-body controller. These results suggest a route to handling bulky objects that are awkward to grip with hands alone, but the abstract does not establish how well the approach generalizes beyond the tested robot and tasks.","question":"How well would the controller perform on different objects, robots, and tasks without additional training or human guidance?","claims":[{"text":"The paper presents a framework in which a humanoid robot uses its body and appendages as contact surfaces to grasp and manipulate objects.","excerpt":"uses the robot's body and appendages as contact surfaces to grasp and manipulate objects.","support":"source-stated"},{"text":"The authors describe training in simulation and report transfer to a real Unitree G1 without additional real-world training.","excerpt":"We show that the distilled policy transfers zero-shot to a real Unitree G1","support":"source-stated"},{"text":"The reported real-robot demonstrations include repositioning a basketball and a box, plus carrying a box from a shelf to a table and placing it.","excerpt":"repositions a basketball and a cardboard box to regions around the torso and completes a long-horizon task that stows a box under one arm, carries it from a shelf to a table, and places it.","support":"source-stated"},{"text":"The controller used coarse motion references from a human teleoperator, so the demonstrations were not described as fully autonomous.","excerpt":"The controller operates interactively, steered by coarse motion references from a human teleoperator.","support":"source-stated"},{"text":"Across the reported real-world tasks, the authors say the distilled controller completed 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","excerpt":"Across real-world whole-body manipulation tasks, the distilled controller completes 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","support":"source-stated"},{"text":"Using the torso and arms as contact surfaces may help with bulky objects that are awkward to handle with hands alone; how broadly this benefit applies is not established by the abstract.","excerpt":"falling short of humans' ability to engage the entire body by bracing and sliding bulky objects against the arms and torso.","support":"interpretation"}],"limitations":["The evidence is the paper's abstract; it does not provide trial counts, task-by-task results, or evaluation details.","The abstract describes one robot model and selected objects and tasks; it does not establish generalization to other robots or settings.","The reported controller relies on coarse movement references from a human teleoperator; the abstract does not show fully autonomous task execution."],"mediaId":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","correctionReason":null,"id":"ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","publicationId":"d9caddec-ba98-4171-8def-2034321b3d80","version":1,"sourceUrl":"https://arxiv.org/abs/2610.22184","sourceTitle":"Whole-Body Dexterity","publisher":"arXiv eess.SY","publishedAt":null,"originalPublished":{"value":"2026-10-15","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-16T04:08:55.924Z","createdAt":"2026-10-18T21:01:44.795Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","kind":"article","url":"https://arxiv.org/abs/2610.22184","sourceUrl":"https://arxiv.org/abs/2610.22184","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-18T21:01:05.397Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2610.22184","checkedAt":"2026-10-18T21:01:05.022Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","confidence":0.87,"provider":"vercel","model":"typesafe-ai/jev","inputHash":"8b7e5a66973b659aa7a0355065c4af9aadf804f479139bf3a371bdcd3e97899c","decidedAt":"2026-10-18T21:01:08.283Z"},"provenance":{"textHash":"d4238b20bb3fed88f5528584c54b13dc7170a881eeaa151a96180297758461ed","observationId":"2d13ca78e9f993a14377aff16dafb4b7510cf10bc57f3560dcb68f97a38cf09c","retrievedAt":"2026-10-18T21:01:05.022Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"7de8cc44-165a-443b-8bc0-b9c0ea50a7d7"},"reviewer":{"model":"gpt-6-luna","runId":"853fdeca-b22a-4894-89b8-579b2b7bf9f8"}},"audit":{"status":"passed","reviewedAt":"2026-10-18T21:01:44.795Z","model":"anthropic/claude-opus-5.5"}},"canonicalUrl":"https://escapemeridian.com/readings/ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","formats":{"html":"/readings/ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","markdown":"/readings/ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de.md","json":"/readings/ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de.json"},"attribution":"Headline and summary by Escape Meridian, prepared by Luna from the attributed source. A source check is not independent confirmation of the source’s claims.","readerContext":{"version":"reader-context-v1","attribution":"Reference context selected by Escape Meridian; separate from source evidence. Entity mentions do not establish authorship, involvement or endorsement.","terms":[],"entities":[{"id":"arxiv","name":"arXiv","kind":"organization","context":"arXiv is a research manuscript repository where authors share papers that may not yet have been peer reviewed.","referenceUrls":["https://arxiv.org/","https://info.arxiv.org/help/api/index.html"],"path":"/entities/arxiv","json":"/entities/arxiv.json","markdown":"/entities/arxiv.md"}]},"withdrawn":false,"references":{"schemaVersion":1,"target":{"type":"reading","id":"ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","revision":"d9caddec-ba98-4171-8def-2034321b3d80","revisionKind":"publication-id","title":"Whole-Body Dexterity lets a humanoid robot use its torso and arms to reposition objects","url":"https://escapemeridian.com/readings/ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de"},"source":{"type":"source","id":"https://arxiv.org/abs/2610.22184","revision":"d4238b20bb3fed88f5528584c54b13dc7170a881eeaa151a96180297758461ed","revisionKind":"sha256-source-text","title":"Whole-Body Dexterity","url":"https://arxiv.org/abs/2610.22184"},"status":"available","items":[],"next":null,"scope":"Explicit citations in public readings and public file attachments to this reading. Shared topics are not citations. Other artifact types need an explicit reference adapter."},"revisions":[{"id":"ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","title":"Whole-Body Dexterity lets a humanoid robot use its torso and arms to reposition objects","argument":"Researchers introduce Whole-Body Dexterity (WBD), a system for manipulating objects by using a humanoid robot’s body and appendages—not just its hands—as contact surfaces. The authors train a controller in simulation on different starting and target grasps, then report that it transferred to a real Unitree G1 without additional real-world training. In their reported tests, it repositioned a basketball and box around the torso and completed a task carrying a box from a shelf to a table and placing it. A human teleoperator supplied coarse movement guidance. Across the real-world tasks they tested, the authors report 90% trial completion for WBD, versus 18% for direct retargeting and 4% for a general-purpose whole-body controller. These results suggest a route to handling bulky objects that are awkward to grip with hands alone, but the abstract does not establish how well the approach generalizes beyond the tested robot and tasks.","topicIds":["intelligence-and-discovery"],"domain":"science","claimType":"attributed reading","limits":"The evidence is the paper's abstract; it does not provide trial counts, task-by-task results, or evaluation details. The abstract describes one robot model and selected objects and tasks; it does not establish generalization to other robots or settings. The reported controller relies on coarse movement references from a human teleoperator; the abstract does not show fully autonomous task execution.","speaker":{"name":"arXiv eess.SY","roleAtSource":"Publisher"},"source":{"title":"Whole-Body Dexterity","publisher":"arXiv eess.SY","format":"Article","url":"https://arxiv.org/abs/2610.22184","publishedAt":"2026-10-15","dateBasis":"Date basis not provided."},"evidence":[{"text":"The paper presents a framework in which a humanoid robot uses its body and appendages as contact surfaces to grasp and manipulate objects.","excerpt":"uses the robot's body and appendages as contact surfaces to grasp and manipulate objects.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The authors describe training in simulation and report transfer to a real Unitree G1 without additional real-world training.","excerpt":"We show that the distilled policy transfers zero-shot to a real Unitree G1","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The reported real-robot demonstrations include repositioning a basketball and a box, plus carrying a box from a shelf to a table and placing it.","excerpt":"repositions a basketball and a cardboard box to regions around the torso and completes a long-horizon task that stows a box under one arm, carries it from a shelf to a table, and places it.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"The controller used coarse motion references from a human teleoperator, so the demonstrations were not described as fully autonomous.","excerpt":"The controller operates interactively, steered by coarse motion references from a human teleoperator.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"Across the reported real-world tasks, the authors say the distilled controller completed 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","excerpt":"Across real-world whole-body manipulation tasks, the distilled controller completes 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","url":"https://arxiv.org/abs/2610.22184","relationship":"Source-stated"},{"text":"Using the torso and arms as contact surfaces may help with bulky objects that are awkward to handle with hands alone; how broadly this benefit applies is not established by the abstract.","excerpt":"falling short of humans' ability to engage the entire body by bracing and sliding bulky objects against the arms and torso.","url":"https://arxiv.org/abs/2610.22184","relationship":"Escape Meridian interpretation"}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2610.22184","checkedAt":"2026-10-18T21:01:05.022Z","items":[],"note":"No suitable media was found by the supported page extraction."},"assumptions":[],"whatWouldChangeView":{"attribution":"Open question","text":"How well would the controller perform on different objects, robots, and tasks without additional training or human guidance?"},"editorialComparisons":[],"live":{"title":"Whole-Body Dexterity lets a humanoid robot use its torso and arms to reposition objects","summary":"Researchers introduce Whole-Body Dexterity (WBD), a system for manipulating objects by using a humanoid robot’s body and appendages—not just its hands—as contact surfaces. The authors train a controller in simulation on different starting and target grasps, then report that it transferred to a real Unitree G1 without additional real-world training. In their reported tests, it repositioned a basketball and box around the torso and completed a task carrying a box from a shelf to a table and placing it. A human teleoperator supplied coarse movement guidance. Across the real-world tasks they tested, the authors report 90% trial completion for WBD, versus 18% for direct retargeting and 4% for a general-purpose whole-body controller. These results suggest a route to handling bulky objects that are awkward to grip with hands alone, but the abstract does not establish how well the approach generalizes beyond the tested robot and tasks.","question":"How well would the controller perform on different objects, robots, and tasks without additional training or human guidance?","claims":[{"text":"The paper presents a framework in which a humanoid robot uses its body and appendages as contact surfaces to grasp and manipulate objects.","excerpt":"uses the robot's body and appendages as contact surfaces to grasp and manipulate objects.","support":"source-stated"},{"text":"The authors describe training in simulation and report transfer to a real Unitree G1 without additional real-world training.","excerpt":"We show that the distilled policy transfers zero-shot to a real Unitree G1","support":"source-stated"},{"text":"The reported real-robot demonstrations include repositioning a basketball and a box, plus carrying a box from a shelf to a table and placing it.","excerpt":"repositions a basketball and a cardboard box to regions around the torso and completes a long-horizon task that stows a box under one arm, carries it from a shelf to a table, and places it.","support":"source-stated"},{"text":"The controller used coarse motion references from a human teleoperator, so the demonstrations were not described as fully autonomous.","excerpt":"The controller operates interactively, steered by coarse motion references from a human teleoperator.","support":"source-stated"},{"text":"Across the reported real-world tasks, the authors say the distilled controller completed 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","excerpt":"Across real-world whole-body manipulation tasks, the distilled controller completes 90% of trials, compared with 18% for direct retargeting and 4% for a generalist whole-body controller.","support":"source-stated"},{"text":"Using the torso and arms as contact surfaces may help with bulky objects that are awkward to handle with hands alone; how broadly this benefit applies is not established by the abstract.","excerpt":"falling short of humans' ability to engage the entire body by bracing and sliding bulky objects against the arms and torso.","support":"interpretation"}],"limitations":["The evidence is the paper's abstract; it does not provide trial counts, task-by-task results, or evaluation details.","The abstract describes one robot model and selected objects and tasks; it does not establish generalization to other robots or settings.","The reported controller relies on coarse movement references from a human teleoperator; the abstract does not show fully autonomous task execution."],"mediaId":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","correctionReason":null,"id":"ac1f4289e350a9ebd7e773995f273219c49c3fd8457d0e77b0353a666c3505de","publicationId":"d9caddec-ba98-4171-8def-2034321b3d80","version":1,"sourceUrl":"https://arxiv.org/abs/2610.22184","sourceTitle":"Whole-Body Dexterity","publisher":"arXiv eess.SY","publishedAt":null,"originalPublished":{"value":"2026-10-15","precision":"day","basis":"publisher_citation_date"},"firstObservedAt":"2026-10-16T04:08:55.924Z","createdAt":"2026-10-18T21:01:44.795Z","historical":true,"entityIds":["arxiv"],"topicIds":["intelligence-and-discovery"],"media":[{"id":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","kind":"article","url":"https://arxiv.org/abs/2610.22184","sourceUrl":"https://arxiv.org/abs/2610.22184","label":"Read original","provenance":"source_link","verification":{"status":"verified","checkedAt":"2026-10-18T21:01:05.397Z","method":"acquired_source","contentType":"text/html"},"recommendationReason":"Recommended by Jev for understanding this source; checked by Luna."}],"sourceMedia":{"status":"not_found","sourceUrl":"https://arxiv.org/abs/2610.22184","checkedAt":"2026-10-18T21:01:05.022Z","items":[],"note":"No suitable media was found by the supported page extraction."},"editorialDecision":{"outcome":"accept","reasonCodes":["material_development"],"rubricVersion":"stream-admission-v5","mediaId":"78ca6a295115580c17efa700ffc58c096945935bde5fef1075f0a90fbbf3cd5b","confidence":0.87,"provider":"vercel","model":"typesafe-ai/jev","inputHash":"8b7e5a66973b659aa7a0355065c4af9aadf804f479139bf3a371bdcd3e97899c","decidedAt":"2026-10-18T21:01:08.283Z"},"provenance":{"textHash":"d4238b20bb3fed88f5528584c54b13dc7170a881eeaa151a96180297758461ed","observationId":"2d13ca78e9f993a14377aff16dafb4b7510cf10bc57f3560dcb68f97a38cf09c","retrievedAt":"2026-10-18T21:01:05.022Z","kind":"publisher-page","preparer":{"model":"gpt-6-luna","runId":"7de8cc44-165a-443b-8bc0-b9c0ea50a7d7"},"reviewer":{"model":"gpt-6-luna","runId":"853fdeca-b22a-4894-89b8-579b2b7bf9f8"}},"audit":{"status":"passed","reviewedAt":"2026-10-18T21:01:44.795Z","model":"anthropic/claude-opus-5.5"}}}],"audits":[{"publicationId":"d9caddec-ba98-4171-8def-2034321b3d80","verdict":"passed","findings":[{"dimension":"attribution","status":"pass","explanation":"The draft attributes the framework and results to the paper's authors, using phrases such as 'the authors report.' It does not present the success rates as independently verified. This matches the arXiv abstract page and its listed authors.","excerpt":"We introduce Whole-Body Dexterity (WBD), a general framework that uses the robot's body and appendages as contact surfaces"},{"dimension":"summary","status":"pass","explanation":"The title, summary and question match the abstract: body-contact manipulation, simulation training, zero-shot transfer to a Unitree G1, the listed tasks, teleoperator guidance and the 90/18/4% results. Zero-shot transfer is explained in plain language. The baseline names are not explained, but they do not stop a general reader from following the development. The question raises a real open issue about generalization.","excerpt":"The controller operates interactively, steered by coarse motion references from a human teleoperator."},{"dimension":"date","status":"pass","explanation":"The originalPublished date of 2026-10-15 matches the v1 submission date on the arXiv page. It is not the later feed date.","excerpt":"[v1]\nThu, 15 Oct 2026 16:40:48 UTC (27,224 KB)"},{"dimension":"support","status":"pass","explanation":"Each factual claim has a verbatim excerpt behind it. The point about bulky objects is labeled as interpretation, worded conditionally, and hedged about how far it generalizes. The limitations correctly note that the abstract gives no trial counts, covers one robot, and relies on human guidance.","excerpt":"falling short of humans' ability to engage the entire body by bracing and sliding bulky objects against the arms and torso."},{"dimension":"media","status":"pass","explanation":"The selected media is the verified direct link to the original arXiv abstract page, which is the source itself.","excerpt":null}],"createdAt":"2026-10-18T21:01:44.795Z","model":"anthropic/claude-opus-5.5"}],"correctionUrl":"https://escapemeridian.com/corrections/"}