A robotic hand reportedly grasps diverse objects without observing contact geometry
A post describes a “see to reach, feel to grasp” approach: vision guides the palm toward an object, then a blind hand reflex takes over.
TLDR
A post says visual guidance positions a robot’s palm before a blind hand reflex grasps the object and the arm continues. It claims robust grasping across diverse objects without observing contact geometry. The linked paper’s abstract describes studying whether a hand can grasp using proprioception—its sense of position—alone, without visual observation.
Combined views
8.4K
2 Sources, first seen 8h ago
