正在加载视频...

视频加载失败

A good hand can push intelligence development. Introducing Eyesight Hand, equipped with full-hand high-res tactile sensors and proprioceptive actuators. It is compliant, agile, and powerful. Good tactile sensing makes learning more efficient and robust. Shout out to Branden!

23,555 次观看 • 1 年前 •via X (Twitter)

15 条评论

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

For more details please visit

Brian Smith 的头像
Brian Smith1 年前

Very impressive work! Did you try providing the vision only model with torque data? It's obviously not as rich as tactile feedback, but I'm wondering how far it gets you. For tasks like cutting it seems like it might perform a lot better

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

That's a pretty good point. We didn't provide torque, but we would love to compare that in the next project. We are also curious about how far proprioception can take us.

Brian Smith 的头像
Brian Smith1 年前

One more question from reading it again today: do you expect GelSim(ple) to need to be calibrated between different hands? The images produced seem pretty clear, but don’t know how much variation the silicone casting process introduces

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

I don’t think it’s necessary though not validated by experiments. When we look at different sensors their images are almost the same, except that the printed dots are not aligned because they are hand made in lab.

Brian Smith 的头像
Brian Smith1 年前

Makes sense, definitely want to give these touch sensors a try at some point since it feels both accessible and very rich. Are they able to detect light touches? Like would it be able to grab an empty soda can without pushing it around much?

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

Yeah, definitely! I think some ppl showed berry picking with these kind of sensors

Brian Smith 的头像
Brian Smith1 年前

That’s great to hear; I’m very excited to try it out

Steven Uecke 的头像
Steven Uecke1 年前

Amazing concept and implementation for the tactical force sensing.

Craig F. Douglass 的头像
Craig F. Douglass1 年前

This is some fantastic work! Very exciting

Yuzhe Qin 的头像
Yuzhe Qin1 年前

Cool design!

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

Thanks Yuzhe!

Huazhe Harry Xu 的头像
Huazhe Harry Xu1 年前

how did I miss this?? It's so cool, Hao-Shu!

Hao-Shu Fang 的头像
Hao-Shu Fang1 年前

Thanks Huazhe!

Nick CleanCode 的头像
Nick CleanCode1 年前

I'm blown away by the Eyesight Hand's capabilities! What kind of applications do you see for this tech?

相关视频

Sharpa Robotics just dropped a new hand video and the level keeps going up. This is the Sharpa Wave running WM Craftnet on a human scale fivefinger hand with 22 active DoF. The policy combines wrist depth, tactile sensing, proprioception and previous actions. The hand can rotate different objects in-hand, recover after external pushes and continue manipulating objects it was never trained on. The numbers are strong. 175/200 successful real-world rotation trials across 20 objects. A world-model prior trained on 9 objects was transferred to 49 new objects. Fall rate went from 6% to 0.3%. The Wave hardware itself has 22 actuators, up to 20 N fingertip force, 240×240 tactile sensing at up to 180 fps and 0.02 N pressure sensitivity. What caught my attention is the recovery behavior. The fingers keep changing contact points after the object slips or gets pushed instead of replaying the same finger motion. That is the kind of dexterity I want to see more of in robotic hands. According to Sharpa’s current specifications: • DTA tactile sensors on the fingers with a resolution of up to 240 × 240 • Pressure detection • Slip detection • Force change detection • Contact point localization • 6-axis force and torque measurement: Fx, Fy, Fz, Mx, My, Mz • Tactile sensing at up to 180 fps • 20 ms reported latency • Force detection range from 0 to 30 N • Maximum sensor load of 50 N • Sharpa also describes a miniature camera integrated into each fingertip for visuo-tactile sensing.

Techniahqrobot | humanoid robots

13,344 次观看 • 6 天前