January 6, 2026 at 2 PM
Soft robots require compliant structures that can both deform and sense their own shape. In this work, we present a fully 3D-printed cable-driven soft robot made of thermo-plastic polyurethane (TPU) foam with integrated conductive TPU sensors. Both the structural and sensing components are fabricated using additive manufacturing. The foam architecture is parametric, enabling control over cell geometry, density, and thus stiffness, allowing the mechanical properties to be tuned for specific actuation requirements. Embedded conductive TPU regions form a distributed resistive network whose electrical resistance varies with local compression. These variations are used to estimate the internal strain distribution, and an inverse finite element model (FEM) reconstructs the global shape of the robot in real time. The cable actuation system provides controllable deformation scenarios that validate the sensing and reconstruction method experimentally. This work demonstrates that 3D-printed parametric TPU foams with integrated conduc- tive sensing networks can serve simultaneously as structure, actuator interface, and sensor, offering a scalable path toward self-sensing soft robots with streamlined fabrication.
Inria building A