Founder and CEO, Lavoro AI, Inc. Director, roBot Intelligence Group (BIG), CMU Associate Research Professor, Robotics Institute, CMU https://xitter.cf/t.co/dtu23qPP8c
Joined December 2021
- Tweets62
- Following25
- Followers109
- Likes42
"As someone who'd never used robots before, … I got full teleoperation running in an afternoon, with a single command."
On day 1 of her Lavoro internship, CMU junior Reya went from zero robotics experience to YAM arms running on RIO in 3 hrs.
Lavoro, Physical AI Made Easy.
Jean Oh retweeted
we built a flexible infra layer for physical ai and robotics research, check out @PAOKApproves's post! switch between different tabletop arms (single, bimanual) & humanoids, for policy inference, teleop, or data collection. all in python, easy for humans and agents to build on.
Physical AI is more than a model. Missing reusable infrastructure is often the painful part.
At #RSS2026, meet RIO: Flexible Real-Time Robot I/O for Cross-Embodiment Robot Learning (1/8)
Jean Oh retweeted
Heartfelt thanks to the full team behind RIO: Pablo Ortega-Kral, Eliot Xing, Arthur Bucker, Vernon Luk, Junseo Kim, Owen Kwon, Angchen Xie, Nikhil Sobanbabu, Yifu Yuan, Megan Lee, Deepam Ameria, Bhaswanth Ayapilla, Jaycie Bussell, Guanya Shi, Jon Francis, Jean Oh. (8/8)
Jean Oh retweeted
So proud of the team behind RIO. Co-led with @etaoxing, guided by @GuanyaShi, @jonmfranc , and @jeanoh at @CMU_Robotics , with wonderful collaborators across TU Delft, Bosch Center for AI, and @lavorobot , the startup where Jean keeps building on RIO. (7/8)
Jean Oh retweeted
RIO is fully open source. We hope it lowers the barrier to deploying, benchmarking, and iterating on modern robot learning across diverse hardware. Learn more at:
📄 arxiv.org/abs/2605.11564 🔗 robot-i-o.github.io 📰 ri.cmu.edu/robot-io (6/8)
Jean Oh retweeted
And it's fast. RIO hits 130.3 ms observation-to-action latency vs 581.2 ms for LeRobot, 4.5x lower, with sub-millisecond middleware options for high-frequency control. Wide platform support, low latency, real-time throughput. (5/8)
Jean Oh retweeted
We validate across 3 morphologies (single-arm, bimanual, humanoid) and 4 hardware platforms. Using RIO-collected teleop data, we fine-tune SOTA VLAs like π0.5 and GR00T on household tasks: folding, pick-and-place, and bowl scrubbing. (4/8)
Jean Oh retweeted
RIO is a lightweight, open-source Python framework of reusable blocks for control, teleoperation, data collection, and policy deployment. One script per task, each runs on any embodiment. Swapping robots or hardware is a config change, not a rewrite. (3/8)
Jean Oh retweeted
The field is fragmented. Varying morphologies, diverse sensors, and vendor-specific drivers make most stacks platform-specific. ROS is powerful but a heavy lift, and existing frameworks lack flexible abstractions at every layer of the pipeline. (2/8)
Jean Oh retweeted
Physical AI is more than a model. Missing reusable infrastructure is often the painful part.
At #RSS2026, meet RIO: Flexible Real-Time Robot I/O for Cross-Embodiment Robot Learning (1/8)
Jean Oh retweeted
RIO, a new framework from CMU, provides a unified interface for robot control, data collection, teleoperation and AI deployment!
It gives researchers a lightweight, modular foundation for deploying robots quickly on any platform.
ri.cmu.edu/robot-io/
Jean Oh retweeted
Missed CVPR/ICRA? Join us for the 1st Embodied4Arts Workshop @ RSS 2026!
Robotics is more than picking, placing, and chasing higher VLA success rates.
We are looking for creative robotics work that explores how robots can paint, draw, perform, craft, fabricate, compose, collaborate, and participate in artistic practice.
We welcome work on robot painting, embodied music, robotic craft, performance robotics, human-robot co-creation, creative manipulation, deformable materials, embodied generative systems, and anything that asks what robots can create, express, or make with us.
Two tracks:
• Workshop Papers & Demos
• Robot Painting Challenge
Submission deadline: June 24, 2026
Workshop date: July 13, 2026
Location: RSS 2026, Sydney, Australia
Website: embodied4arts.github.io/
OpenReview: openreview.net/group?id=robo…
If your robot does something creative, expressive, strange, beautiful, or materially interesting, we want to see it.
#rss2026 #robotics #embodied
Jean Oh retweeted
How can we get robot hands to “hear” slip and contact through microphones, and react to them?
We’re excited to share VibeAct, an approach that uses piezoelectric microphones embedded in robot fingertips to estimate contact and slip, then learns reactive policies from this tactile feedback!
vibeact.github.io
Jean Oh retweeted
SoftAct suggests that contact geometry and force distribution are essential for bridging extreme embodiment mismatch that cannot be recovered through kinematic imitation alone. It would've been impossible without some amazing collaborators at @CMU_Robotics and Meta Reality Labs! Website for more details, robot rollouts, and experiments: soft-act.github.io
Jean Oh retweeted
With the retargeted demonstrations, we train behavior cloning policies and achieve zero-shot real-world success rates of 95% on light bulb screwing, 85% on cup pouring, and 70% on box reorienting vs. 30%, 35%, and 10% for kinematic retargeting.
Jean Oh retweeted
We know soft robot hands 🐙 can do remarkable things like learn dynamic tasks by leveraging stored elastic energy 🖊️and interact safely with people in physical human-robot interaction tasks 💇♀️. But teaching them dexterous skills often relied on handcrafted primitives or embodiment-specific demonstrations. Standard kinematic retargeting fails, suggesting we need to use different information from the demonstrations. 🧵2/7
Jean Oh retweeted
Excited to share SoftAct, a framework for retargeting human manipulation demos to soft robot hands using explicit contact force reasoning! How do you transfer human skill to a hand that looks and moves nothing like yours🐙🖐️? It turns out VR environments can let us capture privileged force interaction demonstrations to help. 🧵1/7
Jean Oh retweeted
🤖What if a robot could understand hair dynamics well enough to style your hair, just like your favorite barber💈?
🔥Excited to announce DYMO-Hair, a model-based robot hair styling system powered by a generalizable 3D hair dynamics model.
🚀A new step toward robots that can understand and manipulate more complex deformable materials.
New dynamics learning paradigm. Fully synthetic data from a novel lightweight simulator. Zero-shot sim2real transfer.
👉Check it out at: chengyzhao.github.io/DYMOHai…
More details below: [1/N]
Due to a number of requests, the submission deadline for the IEEE RA Magazine Special Issue on the Arts and Robotics has been extended to August 15th. The formatting instructions for both technical and non-technical papers can be found on the website: lnkd.in/eBMQNaFC