Rolled self-sensing artificial muscles: Monolithic PDMS integration of SMA actuators and interdigitated capacitive sensors

This study presents a novel rolled self-sensing artificial muscle that integrates shape memory alloy actuators and interdigitated capacitive sensors within a monolithic polydimethylsiloxane matrix. An innovative 3D printing-based fabrication process is introduced for the co-design of actuator and sensor, allowing for the measurement of deformation during actuation. Experimental tests demonstrate the high signal stability and repeatability of the sensors, with gauge factor up to 0.76 and hysteresis lower than 0.013 pF. The actuators achieve peak forces of 12 N, approximately 100 times their self-weight, and displacements of 2.5 mm, corresponding to 3% strain. Actuation time ranges between 85–90 s and is primarily dependent on thermal effects. The dynamic characterization of the self-sensing artificial muscle incorporates both electromechanical and thermal effects, with a temperature compensation term introduced to correct for temperature-induced capacitance variations during actuation. As a proof of concept, a soft robot is developed, realized with three rolled self-sensing artificial muscles arranged in a parallel configuration. A kinematic model is derived to describe the relationship between the measured displacements of the artificial muscles in the actuation space and the three degrees of freedom pose (roll, pitch, and vertical displacement) in the workspace. The findings highlight the effectiveness of the proposed design and fabrication approach for achieving compact, integrated actuation and sensing in a soft robotic device. The proposed rolled, self-sensing artificial muscle is fabricated by integrating shape-memory alloy actuators and interdigitated capacitive sensors within a monolithic polydimethylsiloxane matrix. To validate the design and fabrication process, a 3-DOF soft robot is constructed that uses three self-sensing actuators arranged in a parallel configuration. • A novel rolled self-sensing artificial muscle is proposed which is realized by integrating shape memory alloy actuators and interdigitated capacitive sensors within a monolithic polydimethylsiloxane matrix. • A novel dynamic characterization of the self-sensing actuator incorporates both electromechanical and thermal effects, with a temperature compensation term introduced to correct for temperature-induced capacitance variations during actuation. • A proof of concept of a soft robot, actuated by three rolled self-sensing artificial muscles arranged in a parallel configuration, validates the proposed design and fabrication.