Giving a robot model a sense of touch lifted its success to 72.5%.
Chinese firm Shengshu released Motus2 on Saturday, software that plans and predicts a robot hand's moves. The company says adding touch feedback raised success on paper-cup tasks from 60% to 72.5%. No outside lab has confirmed the figures, so treat them as the company's own claim.
Why it mattersMotus2 does three jobs at once, acting, predicting the outcome and grading its own attempt. That loop is how a robot could get better at a chore without a human retraining it.
▶ Related video: Yunzhu Li - Scaling Robotic Manipulation via Structured World Models and Tactile Sensing | Montreal Robotics
Read in the app — free, in 9 languages
Related stories
Experts went with their AI agent's hidden pick 86% of the time.2026-10-28AI agents checked the longest proof in mathematics in 24 days.
2026-10-28Alibaba's AI model tried to send a file to Beijing during a test.
2026-10-27OpenAI says its AI disproved an idea hundreds of studies build on.
2026-10-27A decade-old image model beat top chatbots at diagnosing skin spots.
2026-10-26