Pinned Tweet
Force sensing for low-cost robot arms — without adding force sensors.
🚀 Excited to share FACTR 2! 🚀
FACTR 2 enables external torque sensing on low cost arms and uses it to improve policy learning.
w/ @JasonJZLiu @_tonytao_
🧵(1/6)
Steven Oh retweeted
Context isn’t just what you put in a prompt. It’s also what stands between a robot and its goal. Robot policies in everyday environments must generalize not only across objects, but also across the spatial contexts in which they appear.
Excited to share COWBOY 🤠 (CoRL 2026) — our sim-to-real recipe for contextual whole-body manipulation. One generalist grasping policy coordinates base, arm, dexterous hand, and active vision end-to-end, adapting its motion and grasps to object and scene geometry. Trained in simulation, deployed zero-shot in the wild.
w/ @JasonJZLiu @LiuMuxinl
🧵(1/N)
Steven Oh retweeted
Introducing PACT: we co-designed a 17-servo anthropomorphic hand and its retargeter together, so teleop demos for imitation learning track the human hand much more closely.
3.9 mm keypoint RMSE on human grasps (vs. 12.8 mm for LEAP Hand).
📄 arxiv.org/abs/2610.18443
Steven Oh retweeted
Coding agents are emerging as general robot policies, but they are still inefficient (e.g., requiring long real-world exploration), especially in mobile manipulation.
Humans become efficient by watching others.
Can coding agents learn from passive human videos, just like we do?
Introducing MoCaP: Mobile Code-as-Policies for Zero-Shot Mobile Manipulation from Passive Human Videos
↓🧵
Steven Oh retweeted
Can we make dexterous robots compliant without sacrificing precise control?
Excited to introduce FOMO-Dex: Force-Modulated Dexterity! 🖐️
FOMO-Dex brings force-aware compliance to dexterous teleoperation, enabling robots to pick up thin cards and chips 🃏, cut paper 🔪, turn nuts 🔩, and use screwdrivers 🪛.
These force-aware demonstrations also train better autonomous policies!
🎥 Videos & paper: yingyuan0414.github.io/fomo-…
🧵👇 (1/8)
Steven Oh retweeted
🤖 For parallel-jaw grippers, high-level plans can already be executed with mature primitives such as grasping. What about dexterous hands?
Introducing GOTT (General Object Trajectory Tracking): object trajectories to dexterous execution with a reusable cross-embodiment contact primitive.
The same primitive supports diverse real-world tasks including sweeping, pouring, whiteboard erasing, and hammering.
With one checkpoint, GOTT reaches 90-96% grasp-and-lift success across multiple objects on three real-world arm-hand platforms.
Check videos 👇
Steven Oh retweeted
We’ve all seen the cool “Astra for robotics” demos. But behind the scenes? Slow rollouts, hand-engineered APIs, and/or simulated behaviors that struggle on real robots.
How do we make coding agent actually work in the physical world? 🦾 Introducing SimEX: Simulation-Integrated Robotics AutoResearch 🧵
arxiv.org/abs/2609.38982
Steven Oh retweeted
Can robots remember what contact felt like, and reuse it in new situations? 🧠 ✋
In our new RA-L paper, we present TaMeSo-bot: a robotic system with a soft wrist that retrieves similar past tactile experiences to guide robust object insertion.
↓🧵
Steven Oh retweeted
How can we leverage pre-trained VLMs for long-horizon, bimanual tasks like cooking? 👨🍳🤖
Our #IROS2026 paper, ViLaIn-TAMP, combines VLM reasoning with structured task & motion planning.
With all the Astra hype, this feels even more relevant, especially since we did it w/ 4o!
Steven Oh retweeted
How can robots learn contact-rich tasks, like insertion, with very little data?
Introducing PHASE, a retrieval-based imitation learning framework that efficiently learns insertion skills by reusing prior experience.
We’ll be presenting PHASE at #IROS2026 in Pittsburgh! 🇺🇸
Steven Oh retweeted
We trained a robot to play football.
How? Self-play for 140 years in a virtual World Cup.
Meet the #Messinator:
Steven Oh retweeted
Accepted to #CoRL2026!🥳
Special shoutout to all my co-authors🐐 who worked hard on this one, see you in Austin!
Steven Oh retweeted
We’ve released code for NEXT, our neural external torque estimation method from FACTR 2!
NEXT enables commodity robot arms to estimate external joint torques without dedicated force sensors. We hope this release will be useful to the robotics community.
Code: github.com/philiphan0109/fac…
Website: jasonjzliu.com/factr2
Force sensing for low-cost robot arms — without adding force sensors.
🚀 Excited to share FACTR 2! 🚀
FACTR 2 enables external torque sensing on low cost arms and uses it to improve policy learning.
w/ @JasonJZLiu @_tonytao_
🧵(1/6)
Steven Oh retweeted
VLAモデルを用いたロボット動作デモの映像を公開しました。
同一構造のモデルを各タスク向けに学習することで、食器のような剛体から布のような柔軟物まで幅広く扱えるようになります。
各タスク向けの学習結果を適宜切り替えることで、動画のような複数タスクの連続実行も可能になります！
Steven Oh retweeted
【論文詳細を公開】
IROS 2026に採択された5件の研究成果の詳細を公開しました。
omron.com/sinicx/activity/ne…
See you in Pittsburgh!
#IROS2026
Steven Oh retweeted
[1/N] Large flow policies (e.g., VLAs) have become extremely popular, unlocking scalability. However, they come at a cost: the robot cannot react quickly. π𝐑² enables them to achieve reactive, real-time control.
Steven Oh retweeted
First demo of reproducing force feedback from the FACTR 2 paper
It relies on torque and current sensing which the feetechs don’t handle as well due to large gear ratios so it’s not perfect. Not sure if limits are in hw or we can do better with core changes & more data
Steven Oh retweeted
Can pretrained VLMs become capable VLAs without losing their semantic capabilities?🤖
Excited to introduce CLAP👏 (Causal Language-Action Prediction), a lightweight VLM-to-VLA fine-tuning recipe that requires no action expert, no vocabulary modification, and no architectural change!
🧵👇🏻
Steven Oh retweeted
Thrilled to share CLAP from my internship at @omron_sinicx !
We turn a pretrained VLM into a VLA with zero architectural changes: describe actions in natural language, prepend them to action tokens. 90.8% on LIBERO with a 2B model, <6 hrs of post-training.
omron-sinicx.github.io/clap/
