Classical Indian dance offers new ideas to train robotic hands

Researchers at the University of Maryland, Baltimore County (UMBC) have discovered that classical Indian dance, specifically Bharatanatyam hand gestures called mudras, encodes a richer and more flexible structure of hand movements than typical everyday actions. By analyzing these mudras through the lens of kinematic synergies—coordinated joint movement patterns that simplify complex hand motions—the team identified six key synergies that offer greater dexterity and accuracy compared to those derived from natural hand grasps. Testing these synergies on American Sign Language letters showed that the mudra-based system significantly outperformed the natural grasp system, suggesting that traditional dance movements represent a "superhuman alphabet" of hand control. Building on these findings, the researchers are applying this enhanced movement framework to robotics and physical therapy. Instead of programming robots to replicate individual gestures, they are teaching robotic hands to combine core synergy patterns to generate a wide range of hand shapes, improving robotic dexterity and naturalness of motion. The team has developed a low-cost