Control and Analysis of a Multimodal Twisted-Winching String Actuator with Embodied Sensing

We present a novel actuator that combines a twisted string actuator (TSA) with a winch mechanism. This hybrid actuator overcomes TSA limitations such as overtwisting and limited stroke length while still offering the ability for high force output. By integrating the winch, greater stroke and variable transmission ratios are achieved without risking overtwisting. The developed kinematic model forms the backbone of a conventional feedback gain controller with a gain scheduling scheme. We allocate control effort proportionally, guided by the relationship between actuator modes and the resulting string statics and velocity. Using a conductive string, displacement can be measured as a function of the change of electrical resistance, enabling precise control with real-time feedback without wires and encoders compromising the compliance of soft robots. Experiments validate our model and control strategies, showing improved performance and responsiveness. By eliminating the need for rigid end-effector encoders and bulky transmissions, this actuator enhances capabilities for soft robotics and automation while advancing TSA technology.