Talk title: Harnessing soft robots’ observability with flexible e-skins
Abstract:
Soft robotic systems are witnessing widespread integration across domains such as medical technologies, industrial automation, and field robotics. Their intrinsic structural compliance renders them particularly advantageous for deployment in unstructured and dynamic environments, enabling adaptive morphology and enhanced safety within human-robot collaborative contexts. However, this same mechanical flexibility introduces significant challenges in state estimation and control, primarily due to the absence of standardized, high-fidelity sensing modalities capable of capturing the continuous, high-dimensional deformations characteristic of soft bodies. In this talk, we posit that with the development of appropriate sensing architectures, the inherent continuum mechanics of soft robots can be leveraged to augment system observability. Specifically, the instantaneous, distributed strain fields within soft structures encode rich information about external force distributions and contact interactions, presenting an opportunity for refined disturbance estimation and contact-aware control strategies. To substantiate this, we investigate soft robotic platforms integrated with proprioceptive e-skins subjected to poorly predictable external loading conditions. By exploiting the coupling between configuration-dependent stiffness profiles and the system’s observability, we demonstrate the feasibility of reconstructing exogenous disturbances solely from proprioceptive feedback. These results quantitatively validate how advanced, multi-modal soft sensory systems can unlock the latent potential of soft continuum robots, paving the way for their integration into complex, realistic scenarios that remain challenging for conventional rigid-bodied robotics.