Why Robot Arms Still Cant Handle a Wet Bar of Soap

The Robotics Podcast with Fexingo: Autonomous Systems, Industrial Robots, and Hardware · Jun 22, 2026 · 7m · transcript

Lucas and Luna dive into the surprisingly hard problem of robotic grasping of slippery, deformable objects. They examine the fundamental physics challenge: how standard grippers rely on dry friction, and why wet soap breaks those assumptions. The episode looks at recent work from MIT's CSAIL lab, where researchers used machine learning to predict slip in real time, achieving a 73 percent success rate on wet soap — up from 12 percent with traditional methods. They discuss tactile sensing, the Gelsight sensor, and why even a simple task like picking up a wet bar of soap is a benchmark for…