Stiffness Actuation

Stiffness actuation involves controlling the stiffness of robotic joints, enabling robots to adapt their interaction with the environment and perform dynamic movements more effectively. Current research focuses on developing novel actuator designs, such as bi-stiffness actuation and differential spiral joint mechanisms, that offer improved control over energy transfer timing and stiffness modulation. These advancements are being coupled with machine learning approaches, including convolutional neural networks, to enhance collision detection and response in collaborative robots. The resulting improvements in robot control and safety have significant implications for diverse applications, including underwater robotics and human-robot interaction.