A new study from MIT’s Computer Science and Artificial Intelligence Laboratory (CSAIL) demonstrates a significant advancement in robotic dexterity. Researchers have developed a system that allows a robot hand to manipulate objects with unprecedented fluidity and human-like adaptability. The system combines high-speed vision sensors with a novel AI model that can react to an object’s movement in real-time, enabling the robot to catch, spin, and balance items mid-air. This approach marks a shift from pre-programmed, rigid movements to dynamic, reactive control, potentially enabling robots to perform complex tasks in unpredictable environments like warehouses or homes. The technology could lead to more versatile assistive robots and improve automation in logistics. Read the full article at: https://technologyreview.com/2024/07/18/robotic-hand-dexterity-ai/