Sparsh-X is the culmination of our multi year work on building a touch foundation model SSL pre-trained for multimodal tactile (image, audio, motion, pressure) Work led by @carohiguerarias, @akashshrm02 and will be presented as oral @corl_conf

100s of robots manipulating 1M+ unique objects in dense clutter in real fullfilment centers that ship to customers we are working on multimodal representations, behavior cloning, reinforcement learning, all with the focus on what actually scales in real deployment conditions

Bored of working on toy tasks in the lab? We are solving robot manipulation at massive real world scales with Vulcan at @amazon I am at #CoRL2025 and my team is looking for PhD research interns, postodcs, and scientists https://t.co/cXGrLDIGEi

Flying out for CoRL from Pittsburgh today! It's already exciting to hear robotics conversations around. Looking forward to see both new and familiar faces. I will be presenting Sparsh-X (Oral presentation session 3) and Sparsh-skin (Poster session 3)!

Introducing CLAMP: : a device, dataset, and model that bring large-scale, in-the-wild multimodal haptics to real robots. Haptics / Tactile data is more than just force or surface texture, and capturing this multimodal haptic information can be useful for robot manipulation.

We have completed the SLAM Handbook "From Localization and Mapping to Spatial Intelligence" and released it online: https://t.co/AnKa398nyw . The handbook will be published by Cambridge University Press. [1/n]

We have completed editing the SLAM Handbook "From Localization and Mapping to Spatial Intelligence" and released it. The 3-part handbook will be published by Cambridge University Press. Enjoy reading online for now!

We're hiring interns (and full-times) all year long! Please email me if interested.

I am particularly excited about how Sparsh-X enables easy sim-to-real transfer of proprioception-only dexterous manipulation policies through tactile adaptation and a few real rollouts https://t.co/YvlzXENJu4

Exciting work on multi-purpose tactile representations, enabling precise insertion tasks and more! Led by friends in our FAIR Robotics team @AIatMeta

Sparsh-X is our new 'multi-sensory' rep model showing the power of encoding multiple tactile modalities for dexterity. We saw performance for plug insertion go up to 90% success with all modalities, and higher robustness with sim-to-real policies adapted to use Sparsh-X!

Excited to release AlgoTune!! It's a benchmark and coding agent for optimizing the runtime of numerical code 🚀 https://t.co/bdR630y0dL 📚 https://t.co/vSnV3eUgVs 🤖 https://t.co/krJ7XDrJFA with @OfirPress @ori_press @PatrickKidger @b_stellato @ArmanZharmagam1 & many others 🧵

We demo’d Amazon grasp model at RSS this year. We performed over 600 grasps over one day at roughly 80-90% SR 1. On an open item set (people gave random often adversarial items), 2. In a random scene fully outdoor throughout the day 3. On a new embodiment (different from

So you’ve trained your favorite diffusion/flow based policy, but it’s just not good enough 0-shot. Worry not, in our new work DSRL - we show how to *steer* pre-trained diffusion policies with off-policy RL, improving behavior efficiently enough for direct training in the real

I'll be speaking at the RSS Dexterous Manipulation Workshop tomorrow, discussing our recent work with Atlas!

📢Life is a sequence of bets – and I’ve picked my next: @MyolabAI It’s incredibly ambitious, comes with high risk, & carries unbounded potential. But it’s a version of the #future I deeply believe in. I believe: ➡️AI will align strongly with humanity - coz it maximizes its own

https://t.co/aqJxVOQyII

This was a key feature in enabling DexterityGen, our teleop that can support tasks like using a screw driver Led by @zhaohengyin, now open source

Everyday human data is robotics’ answer to internet-scale tokens. But how can robots learn to feel—just from videos?📹 Introducing FeelTheForce (FTF): force-sensitive manipulation policies learned from natural human interactions🖐️🤖 👉 https://t.co/CZcG87xYn5 1/n

Very interesting work! And great to see self-supervised learning being used for tactile data. This is critical to scaling tactile to the level that vision has scaled.