This Fall, I'm teaching a new "Hands-on Robot Learning" class at @JHUCompSci A full-stack class where students will get SO-101 robot kits, build robots out, collect data, train and deploy learned policies/WAMs/agents/etc. Materials will be posted here: hands-on-robot-learning.gith…

Making it easy to shape tactile sensors for arbitrary surface geometries is really important for making these kind of sensors as ubiquitous as cameras! Where you mount sensors plays a pivotal role in how useful they are for a robot. It is even more true for tactile sensors where contact is required for signal!

What if you could give almost any robot a sense of touch? OpenTac is an open-source toolkit that lets you design custom tactile sensor skins for robot fingers, grippers, hands and even robot feet. Its CAD tool converts a 2D outline or 3D model into fabrication-ready flexible PCBs and 3D-printed sensor covers. The sensors detect both pressure and shear, helping robots understand when they're touching, pushing or slipping against an object. Its adaptive readout reaches 1286 partial updates per second on a 1,024-sensor array which is quite impressive. What would you build if your robot could feel?

here are some new designs for tendon-based hands to arm

Touch sensing for any part of the robot, not just the fingertips. OpenTac designs a tactile skin from a 2D outline or a 3D model and gives you the PCB files plus 3D-printed bumps. They built and tested 31 designs this way. The interesting finding is shear. For plug insertion without vision, adding normal force alone didn't help at all over proprioception. Shear is what made it work. The trick is small bumps that tilt when pushed sideways. The skins are also tough: hammered, thrown and run over by an office chair, and they still work. And a smart readout gets about 1,300 partial frames per second from a 1,024-taxel skin, vs 67 full frames with normal scanning. The repo even ships a skill, so Claude Code can design a sensor for you.

Robotics labs that want touch sensing can now build their own. OpenTac is open source: design a tactile skin for any part of the robot, read it out, and simulate it.

ロボットの「触覚」を、様々な表面に実装できるOpenTac。 3D触覚センサーをロボットの形に合わせて設計・製造、データ取得、異なるセンサー間でのデータ共有などを支援。 ロボットごとに異なる触覚システムを構築する負担が減れば、Physical AIに大きく影響しそう。

Arik Slepyan presents OpenTac, an open ecosystem for 3D tactile sensing on any robot surface: design and fabricate high-density tactile skins, read them out, simulate them and share data across sensors. Credit: @arikslep

Echo the broad sentiment, but with one edit: wrapping flat tactile sensors around curved surfaces is *not* physically impossible -- it just requires clever design

If the barriers to entry for tactile sensing for you are: - "It's too hard -- I need to have mechanical and electronic design expertise", - "Aren't the sensors usually very fragile?", - "I don't know where even to start", - "Assembling and customizing sensors is a lot of work", or - "It's very expensive. Who'd pay $500 for a fingertip sensor that wears out in a week of heavy use?" then you'll want to try OpenTac. @arikslep and team worked extra hard to ensure everything 'just works'

Amazing work by the Johns Hopkins team on tactile sensing and compliant control! Proud to support researchers pushing the boundaries of what’s possible with Physical AI through our Research Grant Program.

The sensors withstand hammer strikes, stomping and office-chair loading without a change in response, and show no false response under a heat gun. The readout gain sets the range, about 0.1–10 N by default (configurable up to 60 N).

We're getting better at teaching robots where things are. The next challenge is teaching them what to do when they make contact. Projects like OpenTac are helping make that feedback more accessible.

This is one of the best examples of robustness I have seen!!! 🤓

That's what you get for shipping a CAD tool and order-ready boards with the paper instead of a "code coming soon"

Wild! 1. We put out a preprint this afternoon. 2. It hasn't even hit arXiv yet, 3. and yet we have someone build a nontrivial design using our method by daisy-chaining multiple OpenTac sensors.

That clip of testing robustness is toptier. 😂

Tactile sensing is one modality that struggles intensely with data scarcity. That's why, I think efforts to unify hardware across the field is very valuable for future research. Very exciting to see this! @arikslep

Whole-body tactile skin with shear force is amazing! Can't wait to try this!

High-resolution, easy-to-make tactile sensors that measure *both normal and shear force* for any robot surface area! This is incredible @arikslep @mangahomanga

Tactile sensor for any body, now open source that you can build and configure.