🚀 Thrilled to present ViSafe, a vision-only airborne collision avoidance system that achieved drone-to-drone avoidance at 144 km/h. In an era of congested airspace and growing autonomy, reliable self-separation is paramount 🧵👇

🦾How can humanoids unlock real strength for heavy-duty loco-manipulation? Meet FALCON🦅: Learning Force-Adaptive Humanoid Loco-Manipulation. 🌐: lecar-lab.github.io/falcon-humanoi… See the details below👇:

Basking in the glow of Commencement 🎓✨ During the 127th Commencement ceremony, a class of 6,741 Carnegie Mellon graduates emerged, ready to solve complex challenges and drive transformative change. Congratulations to the Class of 2025! #TartanProud

🚧 High-quality, diverse data is a major bottleneck in robot learning. 🌎🤲 DexWild offers a fast, scalable way to collect human demonstrations in-the-wild — unlocking real-world generalization for robots! #CMUrobotics

Appleseed labs, a team from Kantor Lab, won the Excellence in Regenerative Agriculture award at the 2025 Farm Robotics Challenge! 🌽🏆👏 Their reforestation robot showed special commitment to environmental sustainability + ecosystem health. Learn more: loom.ly/dw8v3h0

On my way ✈️ to ATL for IEEE ICRA! LeCAR Lab at CMU will present 8 conference papers (including DIAL-MPC as the Best Paper Finalist) and one RA-L paper. Details: lecar-lab.github.io Hope to meet old & new friends and chat about building generalist 🤖 with agility 🚀

Robots need touch for human-like hands to reach the goal of general manipulation. However, approaches today don’t use tactile sensing or use specific architectures per tactile task. Can 1 model improve many tactile tasks? 🌟Introducing Sparsh-skin: tinyurl.com/y935wz5c 1/6

How can robots avoid both known and unseen failures in the open world? Take Jenga: a tower can collapse in countless ways when pulling a block. Even without seeing all those failures, our uncertainty-aware latent safety filter (UNISafe) can reliably prevent failures! (1/9)

Congrats to Jiaoyang Li (Jiaoyang Li) on receiving the NSF CAREER award! Her work, “Environment Optimization for Large-Scale Multi-Agent Coordination: Models, Methods, and Applications” reimagines how environments are designed for robots! Learn more👇 loom.ly/6etJMLc

The 2nd CMU Vision-Language-Autonomy Challenge is now open for registration!🤖The challenge focuses on vision language navigation and scene understanding, with a specific focus on resolving object-centric spatial relations. Check it out/register here! ➡️ai-meets-autonomy.com/cmu-vla-challe…

Congrats to Ioannis Gkioulekas for receiving the 2025 Bodossaki Distinguished Young Scientist award in "Applied Sciences and Technologies: Artificial Intelligence" 🏆 The award recognizes scholars of Greek descent for significant scientific progress! ⬇️ loom.ly/vZ70tv4

Humanoids can run, parkour, dance, and fight nowadays, but can it serve🍺☕️ just like wheeled robots? It could be hard - we found that the default G1 controller yields an average end-effector acceleration ~0.5G (human-level is ~0.05G)! Check-out👇to see how we reduce this

Robots may struggle in cluttered scenes due to challenging occlusions 🤖🚧 RaySt3R is a new foundation model that predicts the full 3D geometry from a single view! 👇👇 #CMUrobotics

Say ahoy to 𝚂𝙰𝙸𝙻𝙾𝚁⛵: a new paradigm of *learning to search* from demonstrations, enabling test-time reasoning about how to recover from mistakes w/o any additional human feedback! 𝚂𝙰𝙸𝙻𝙾𝚁 ⛵ out-performs Diffusion Policies trained via behavioral cloning on 5-10x data!

Introducing UFM, a Unified Flow & Matching model, which for the first time shows that the unification of optical flow and image matching tasks is mutually beneficial and achieves SOTA. Check out UFM’s matching in action below! 👇 🌐 Website: uniflowmatch.github.io 🧵👇

Shortcut models enable scaling offline RL, both at train-time at test-time! We beat so many other algorithms on so many tasks we had to stick most of the results in the appendix 😅. Very proud of Nicolas Espinosa Dice for spearheading this project, check out his thread!