Latest / The Robotics Podcast with Fexingo: Autonomous Systems, Industrial Robots, and Hardware
Why Robot Arms Still Can't Grasp a Soft Ball
Lucas and Luna dive into one of robotics' most stubborn problems: grasping deformable objects. They use the specific case of a tennis ball — something a human hand can grab without thinking — to illustrate why even the most advanced robot arms fail at it. They break down the physics of compliance, the limits of current force-torque sensors, and why deep learning alone hasn't solved the problem. They also touch on a recent paper from MIT's CSAIL that proposes a tactile-sensing approach using GelSight sensors, and why it's still not reliable enough for a factory floor. Along…
The skinny
The skinny isn't ready yet — notes appear once the transcript is processed.