Compliant Actuator

Compliant actuators are robotic components designed to mimic the flexibility and adaptability of biological muscles, prioritizing safe and effective interaction with unpredictable environments. Current research emphasizes developing actuators with variable stiffness, often using series elastic actuators or novel designs inspired by biological systems, and integrating them with advanced control algorithms like deep reinforcement learning for optimal performance and design. This work is driven by the need for safer and more dexterous robots in applications ranging from assistive robotics and human-robot collaboration to aerial robots navigating cluttered spaces, improving both robot performance and human-robot interaction.

Papers