Nathan Boyd @ CoRL 🐙 (@servo_boyd) header
Nathan Boyd @ CoRL 🐙 (@servo_boyd) profile picture

Disney Imagineering R&D Lead | Top 0.1% of Chipotle goers | Views are my own

Joined June 2014
Nathan Boyd @ CoRL 🐙 retweeted
There are many papers solving motion tracking nowadays and we take a step back and look at the motions to track and question: Can legged robots imitate human motion when their kinodynamic properties demand 𝗱𝗶𝗳𝗳𝗲𝗿𝗲𝗻𝘁 𝘁𝗶𝗺𝗶𝗻𝗴? To answer this, we are happy to share DTMR!
4
9
1
69
3,298
Nathan Boyd @ CoRL 🐙 retweeted
it was only a matter of time. if you do have a bunch of GPUs though, I got training down to 27 seconds on 4x RTX 6000 Pro. the whole pipeline, including retargeting, takes under a minute.
Can a laptop train a humanoid robot controller? Training a humanoid motion controller used to mean firing up a cluster of NVIDIA GPUs. Now you can train in 22 minutes on a MacBook Pro. Project page: air23zj.github.io/macmimic/
5
4
2
163
9,080
Nathan Boyd @ CoRL 🐙 retweeted
there was lots of speculation at CoRL about what Skild was actually doing. I don’t think we should be so surprised 😉
Steel 0.2 is out: • 81 simulation environment examples, interactive docs, and migration instructions for your coding agent. • Magnetic contacts that latch, hold, and release. • Closed-loop joints for grippers and parallel robots.
1
2
2
83
12,349
Nathan Boyd @ CoRL 🐙 retweeted
In a collaboration with Oxford, SNU, and Pisa, we developed a great legged-robot state estimator, from proprioception only, by being careful about when and how often to use contact events. Check out the results (and code) at dynamic.robots.ox.ac.uk/proj…
3
18
1
147
9,019
Nathan Boyd @ CoRL 🐙 retweeted
NVIDIA just dropped 10,000 OpenUSD objects on Hugging Face 🤯 → a microwave with a hinged door → a dresser with sliding drawers → a forklift with a lifting mast → a wheelchair with spinning wheels 1,000 categories, 8,222 with joints, real mass and collision. Ready for Isaac Sim, Apache 2.0. ⬇️ huggingface.co/datasets/nvid…
21
99
12
1,174
67,697
Nathan Boyd @ CoRL 🐙 retweeted
Qiayuan (@qiayuanliao) and I wrote about modeling the Unitree G1’s closed-loop ankle linkage in @mujoco. We look at how the linkage changes the ankle’s motion and dynamics, and what it takes to simulate it reliably. We spent quite a bit of time on the blog post making interactive demos and shareable models, and we hope you find it useful! kzakka.com/closing-the-loop/
15
69
6
403
24,506
Nathan Boyd @ CoRL 🐙 retweeted
We're excited to release the first open-source VIO to beat Aria on LaMAria! Industry level SLAM for everybody will power tons of applications, can't wait to see what people build :) So how did we beat Aria's SLAM? (1/4)
when LaMAria was introduced a year ago, Meta's closed-source Aria SLAM was far ahead of academic systems. one year later, we've surpassed it in 3/5 categories: medium, long, and low light. here's what closing that gap looks like:
32
80
11
858
435,388
Nathan Boyd @ CoRL 🐙 retweeted
CoRL was a blast, met a ton of great people! To celebrate, I taught a G1 to ride a mechanical bull the Texan way 🤠🐂 Inspired by @the_real_btaba's demo at the MuJoCo meetup, but this time in mjlab and holding a rope! It's already better at it than I am...
5
8
2
102
6,570
Nathan Boyd @ CoRL 🐙 retweeted
Slides of my talk today: drive.google.com/file/d/1etd…
Join us this afternoon 2pm at Agentic Robotics Workshop at CoRL! Great speakers from UC Berkeley @Ken_Goldberg , CMU @GuanyaShi , NVIDIA, Gemini Robotics, Dexmate, and Tencent! And enjoy cool agentic demos! Location: JW Room 203-204 Demo: Griffin Room (2nd floor)
4
20
2
186
11,892
Nathan Boyd @ CoRL 🐙 retweeted
pendulum has fully swung back from where it was last year (“scale up data collection, behavior clone large models”). everyone is finally seeing the shining light of RL!
Imitation learning plateaus with more data. Additional benefit of collecting more data stops improving performance -- hence, Kay was convinced they needed reinforcement learning.
5
8
3
93
14,979
Nathan Boyd @ CoRL 🐙 retweeted
Apptronik's Apollo humanoid could be heading for a sizeable warehouse rollout GXO has signed a letter of intent covering up to 1,000 of the robots over the next four years although this is subject to successful trials and initial commercial deployment targeted for late 2027 Earlier this week, Apollo 2 was demonstrated using Google's new Gemini Robotics 2.5 for autonomous parts kitting in a Mercedes Benz factory setting, which they with a 96% task success rate These developments are building on longer running partnerships these companies have with Apptronik
5
4
4
42
5,656
Nathan Boyd @ CoRL 🐙 retweeted
"Every safety milestone we achieve is a door that opens for us. Cooperative safety unlocks environments and use cases we couldn't enter before. Those, in turn, unlock the operational experience required to continuously evolve our safety technology and achieve the next tier of safety concepts and certification." — Pras Velagapudi and Kevin Reese Read more: agilityrobotics.com/content/… #MadeForWork #Agility #Digit
18
19
2
102
7,110
YEE HAW WITH US AT CORL THURSDAY NIGHT 🤠.
🤖for folks at CoRL: tomorrow (Thursday) mechanical bull riding starting at 7p at Bull and Bowl, going out after that. ending the conference Texas style! 🤠 cowboy hats may or may not be provided 👁️👄👁️
2
10
1,529
Nathan Boyd @ CoRL 🐙 retweeted
We're experimenting with goal-conditioned locomotion in simulation. We give it a goal and let it figure out how to get there. The results are somewhere between locomotion, navigation, and parkour. Ayg has apparently decided to become a mountain goat 🐐
Teaching a robot to recover on a flat floor is one thing. The real world makes it harder: nobody guarantees that when the robot falls, it will land nicely on a flat surface. Stairs, slopes, edges, uneven contacts, weird orientations... the recovery policy still has to get the robot back on its feet. So that’s what we’re testing now. Still lots of work ahead, but the recovery policy is holding up surprisingly well 😃
3
7
84
3,221
Nathan Boyd @ CoRL 🐙 retweeted
Deploy the artifact, not the framework. How we run ProtoMotions policies on robots: the whole pipeline exported as one graph, a small C++ loop with TensorRT, the same binary in MuJoCo and on the robot. 1.3 ms per step.
4
30
1,359
Nathan Boyd @ CoRL 🐙 retweeted
CoRL Keynote. Hilton Austin. Monday Nov 9, 1:45 pm. 2 days countdown. Something extraordinary.
80
121
64
1,361
241,629
Nathan Boyd @ CoRL 🐙 retweeted
Did this because I think AME-2 deserves a complete verification on humanoids. It might be a bit late tho. Some sim perf: collision avoidance on sparse stepping stones, foot-only climbing of 70 cm obstacles, 1.2 m gap traversal (can be larger I guess)... github update soon.
1
10
63
2,651
Nathan Boyd @ CoRL 🐙 retweeted
Are you attending #CORL2026? The #Apptronik team is looking forward to connecting with top talent in robotics and embodied AI at our “Fueling the Future: Apptronik CoRL Networking Lunch” on Wednesday 11/11 in Austin! Whether you’re curious about what we've been building, interested in career opportunities, or just looking to talk robotics with fellow researchers, we’d love to connect. To join us, please submit your RSVP below. Space is limited! luma.com/97z1wrhu
2
4
1
38
4,313
I will be attending @corl_conf next week! If anyone wants to talk RL and Robotics for Humanoids or learn more about the work we are doing at Disney Imagineering R&D for walking characters shoot me a DM or comment!
11
49
3,020
Nathan Boyd @ CoRL 🐙 retweeted
Humanoid retargeting and IK in microseconds [no GPU required] Orders of magnitude speedups, with perfect contact handling: accelerate your robots today at github.com/StanfordASL/frax w/ @drmapavone
10
56
8
558
34,281