Designing Functional Architectures for Soft Robotics

Abstract

Soft robotics enables compliant and safe interaction in complex and continuum environments; however, the integration of reliable, electronics-free sensing remains a major challenge for fully pneumatic robotic systems. Current pneumatic strain gauges predominantly employ simple microchannel geometries, which limit pressure sensitivity, structural efficiency, and spatial sensing capability. This thesis investigates the use of gyroid-based Triply Periodic Minimal Surface (TPMS) architectures as internal structures for pneumatic tactile sensing, with the aim of overcoming limitations in conventional pressure sensing approaches. Parametric gyroid geometries were designed using computational methods, and their mechanical response under loading was initially evaluated through finite element analysis (FEA). In addition, an experimental proof-of-concept setup was developed, integrating multiple pressure gauges within a pneumatic network to capture real-time pressure variations under different touch-points. A soft pneumatic sensing system was developed and experimentally evaluated to assess pressure sensitivity and response behavior under various loads. The results indicate that the gyroid structure enables measurable and distinguishable pressure changes across multiple sensing touch points, when changes in internal flow due to structural deformation. While variations in sensor response were observed depending on touch-point location and loading conditions. These findings demonstrate that TPMS architectures can improve pneumatic sensing response by overcoming the drawbacks such as leakage, ballooning effect, and the global pressure response in the traditional pressure sensors.