Lei Jun-Backed MIT Team Uses Cameras for Robot Touch, Raises Hundreds of Millions

The MIT-linked team, which previously attracted investment from Lei Jun, is taking a different route from conventional strain-gauge or capacitive tactile sensors. Its approach points a camera at a soft, deformable contact surface; when the robot touches an object, the surface deforms, and vision algorithms reconstruct contact geometry, force distribution and slip in real time. That gives robotic grippers and hands a sense of touch without dense sensor arrays. The company has now raised hundreds of millions of yuan to scale the technology, according to Leaderobot. The funding signals growing investor interest in embodied AI components, where reliable tactile feedback is seen as a bottleneck for dexterous manipulation, fragile-object handling and safer human-robot interaction. The team says the camera-based design could make tactile sensing cheaper and easier to deploy across robot platforms.