VQV Signal

OPEN SOURCE WATCH TECHNICAL

Daimon Robotics says touch data lifted a peg-in-hole task from just over 10% to 80%+ success

According to a post on X by @DaimonRobotics, CEO Jianghua Duan argues in a new podcast episode that robots can learn with nearly 10x less data when training includes touch. He cites a peg-in-hole assembly task whose success rate rose from just over 10% to more than 80%. The post also says Daimon has open-sourced 10,000 hours of tactile-enabled manipulation data, though the snippet is cut off at that point.

Source: Scobleizer News Pulse (X) · x.com Found via Robert Scoble’s Scobleizer News Pulse · a post on X, not independently verified Published 2026-11-21T05:42:00+00:00 Detected 2026-11-21T16:07:43+00:00
View original source

According to a post on X by @DaimonRobotics, CEO Jianghua Duan argues in a new podcast episode that robots can learn with nearly 10x less data when training includes touch. He cites a peg-in-hole assembly task whose success rate rose from just over 10% to more than 80%. The post also says Daimon has open-sourced 10,000 hours of tactile-enabled manipulation data, though the snippet is cut off at that point.

AI-assisted summary based on the listed source.

In a new podcast episode, Daimon Robotics CEO Jianghua Duan argues robots can learn with nearly 10x less data when training includes touch, citing a peg-in-hole assembly task that rose from just over 10% to more than 80% success. Daimon has open-sourced 10,000 hours of tactile-enabled manipulation dat...

Robotics teams working on assembly and manipulation may find the claim relevant to how they collect training data, especially if tactile sensing is part of their hardware. The figures come from the company and its CEO, so they have not been independently checked here. Watch for the open-sourced dataset and for third-party results testing whether the gains hold on other tasks.

Signal Strength 83% Technical label WATCH Public Interest 35 Category OPEN SOURCE Reader Depth TECHNICAL

Signal Strength reflects source quality, relevance, freshness and evidence. Public Interest helps organize discovery; it is not proof of truth.

Public Interest components
Recognizable Entity Score 0 Practical Impact Score 0 Novelty Interest Score 94 Consequence Score 22 Curiosity Score 84 Shareability Score 45

VQV surfaced this signal because it is recent, relevant to Robotics, connected to Scobleizer News Pulse (X).