A Soft Gripper of Dielectric Elastomer Actuator Controlled by an LSTM-Based Machine Learning Mode

Soft robotics has gained considerable attention due to their suitability for applications requiring gentle, energy-efficient, and adaptive interactions. A key challenge in soft robotics and smart materials domain are enabling real-time control to enhance the functionality and responsiveness of soft robotic systems. Our study introduces an innovative system that integrates a Dielectric Elastomer Actuator (DEA)-based soft gripper with a machine learning (ML)-based control model. The system utilizes motion data, formerly acquired through controlling signals, to train the ML model, enabling real-time decision-making and adaptive manipulation via a user-friendly application interface.