
research intern @amazon | robot learning research @UCSanDiego
San Francisco, CA
Joined April 2020
- Tweets213
- Following420
- Followers536
- Likes455
Pinned Tweet
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
Update: schedule changed and I will be in poster session #2 at 4:30-5:00pm
Excited to share that Whole-Body Dexterity was accepted as a poster at the upcoming CoRL Dexterous Manipulation: Full-Stack Robotics on Learning and Design workshop! 🦾
Looking forward to chatting during the poster session next week!
📍Thursday, Nov 12, 3:30-4:00 PM
2
279
Conor Mc Gartoll retweeted
Come chat about RoboKit at #CoRL2026 today! 🤖
🌟 Spotlight S6: 8:30–9:00 AM at Hilton Austin
🪧 Poster #105: 10:30 AM–12:00 PM at JW Marriott
Check out the code: github.com/openrobokit/robok…
Excited to share OpenRoboKit, our series of open-source robotics projects! 🤖
- RoboKit for fast, scalable optimization & simulation github.com/openrobokit/robok…
Our RoboKit paper has been accepted for a spotlight at #CORL2026
- RetargetingKit for motion transfer & teleoperation github.com/openrobokit/retar…
- GraspKit for end-to-end grasping github.com/openrobokit/grasp…
Website: openrobokit.com
Check the video (sound on 🎵) for a 16-hour grasping time-lapse in beautiful San Diego, plus a quick look at dexterous retargeting and robot simulation.
Conor Mc Gartoll retweeted
Find it. Reach it. Grasp it. Easy for us, hard for a humanoid.
Introducing HuG (#CoRL2026): Learning Humanoid Whole-Body Affordance-Aware Grasping with Active Perception.
The robot looks around for a half-hidden object, squats or stretches its whole body to reach it, and grasps. Zero real demos, zero-shot to unseen objects and everyday scenes, from low tables to high shelves.
🌐 xuanbinpeng.com/hug/
Excited to share that Whole-Body Dexterity was accepted as a poster at the upcoming CoRL Dexterous Manipulation: Full-Stack Robotics on Learning and Design workshop! 🦾
Looking forward to chatting during the poster session next week!
📍Thursday, Nov 12, 3:30-4:00 PM
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
Conor Mc Gartoll retweeted
Happy to share Catch It All!
We catch completely unseen, random objects with lightning speed, safe reactions ⚡
Thank you to @arth_shukla, @ncklashansen, @HuazeLiu, @ABuynitsky, @darin_djapri
💻 Code & results available at:
arth-shukla.github.io/catch-…
Really excited to share Catch It All!, my first paper as last author! In this work led by @Kehlani_Fay @arth_shukla we develop a system for catching diverse objects using a single egocentric camera 🧵👇
Conor Mc Gartoll retweeted
Really excited to share Catch It All!, my first paper as last author! In this work led by @Kehlani_Fay @arth_shukla we develop a system for catching diverse objects using a single egocentric camera 🧵👇
Extremely impressive forceful wbc from @dangan1693 and Arturo. MPC force reasoning makes training RL policies quite sample efficient. Glad to be involved in this project!
Can a single humanoid controller 🏋️ power clean, 🛒 pull heavy carts, 📦 carry bulky objects against its body, and 🚶 stay balanced while walking?
Introducing HULK 💪: Learning Whole-Body Forceful Loco-Manipulation for Humanoids.
I co-led with Arturo Flores Alvarez during our internships at Amazon.
🌐 hulk-wbc.com
More below 🧵👇
13
1,030
Conor Mc Gartoll retweeted
Blind robot policies indeed can get quite far on capabilities.
Two key enablers broadly emerging are: (1) proxy tactile input of error between commanded and achieved joint angles and (2) contact exploring behavior emerging from blind RL experts.
True tactile should help us go even further.
One thing I was surprised by this project was how far you can go with proprioception only. I was a little skeptical until we got a full loop in real for the first time lol
There are a lot more works recently on blind policies from @aditya_bhatt, @BronarsToni, @younghyo_park and others that show some really cool results! Looking forward to more research in this space and revisiting fusing different modalities.
Conor Mc Gartoll retweeted
Humanoids today mostly manipulate objects with their hands. Humans use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the entire humanoid body into one big dexterous manipulator. Work led by @cmcgartoll, @dangann66, @pthangeda_.
🌐 wholebodydex.com
Conor Mc Gartoll retweeted
So proud of this one! 🤖 Dexterity we usually expect from a multi-finger hand (rolling, sliding, handoffs), now across the whole humanoid body.
Amazing work @cmcgartoll @dangan1693 @pthangeda_ @arjun__gupta @carohiguerarias @mukadammh @ManikantanNambi!
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
Really great and clean agentic toolset from @jianglong_ye, @kaylee_keyi, @gengchen, and @Isabella__Liu. So awesome to see it released!
We introduce RoboSkills: a minimal, low-cost skillset for agentic robotics 🤖
It supports diverse robotics workflows, such as learning cat’s cradle 🧶 from a human video (Real2Sim) and transferring multi-person motion to characters and humanoids 💃 (Motion Generation).
We also support Auto Research, Agent as Policy, and Agentic Simulation. With suitable tools and workflows, we enable cheaper models to solve complex tasks.
GitHub: github.com/openrobokit/robos…
Website: openrobokit.com/roboskills
Watch the video for more examples.
Conor Mc Gartoll retweeted
A lot of humans' robustness and generality (as embodied agents) comes from this sort of capability
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
Conor Mc Gartoll retweeted
Really interesting sim2real approach, and awesome videos!
The whole thread is great, I really like the way feasible holding modes are generated and used to train (via RL + distillation) a whole-body controller optimized for a range of downstream tasks!
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
Conor Mc Gartoll retweeted
We introduce RoboSkills: a minimal, low-cost skillset for agentic robotics 🤖
It supports diverse robotics workflows, such as learning cat’s cradle 🧶 from a human video (Real2Sim) and transferring multi-person motion to characters and humanoids 💃 (Motion Generation).
We also support Auto Research, Agent as Policy, and Agentic Simulation. With suitable tools and workflows, we enable cheaper models to solve complex tasks.
GitHub: github.com/openrobokit/robos…
Website: openrobokit.com/roboskills
Watch the video for more examples.
Conor Mc Gartoll retweeted
Humanoids today mostly manipulate objects with their hands. As humans, we use much more than our hands: we brace boxes against our chests, tuck objects under our arms, and roll balls along our forearms.
Introducing Whole-Body Dexterity (WBD), turning the humanoid body into one large dexterous manipulator.
wholebodydex.com 🧵👇
In zero-shot deployment on a real Unitree G1, WBD succeeds in 90% of trials, compared with 18% for direct retargeting (GMR) and 4% for a generalist whole-body controller (SONIC).
Underarm basketball stow: 9/10 round trips, compared with 2/10 and 0/10.
Shelf box pick → stow → carry → place: 7/10, compared with 1/10 and 0/10.
1
3
291
📜 arxiv.org/abs/2610.22184
🌐 wholebodydex.com (check out interactive simulations!)
Work done during my internship @ Amazon and excited to present it at the CORL Dexterous Manipulation Workshop!
Huge shoutout to my co-authors @dangan1693 and @pthangeda_ (for the many late nights we spent working on this), as well as the entire team @arjun__gupta, @carohiguerarias, @NimaFazeli7, @mukadammh, and @ManikantanNambi
A couple of fun sim-to-real examples made possible by WBD: carrying two boxes at once and walking while moving a basketball around the body.
2
300












