Physically intelligent soft antennae enhance tactile perception by active touch

Abstract

Soft robotic sensors today struggle to interpret complex tactile scenes without incurring significant computational costs. Inspired by insect antennae--compliant, distributed sensors that efficiently process tactile information through physical intelligence--we investigated whether mechanical design and active touch sensing strategies could enhance robotic tactile feature perception. We hypothesized that insect-inspired antenna dynamics, specifically flexural stiffness gradients and active touch speed, could simplify tactile classification. Using a sim-to-real framework that bridges bioinspired computational models with a multi-link soft robot antenna, we introduce the notion of tactile fields--spatiotemporal representations of tactile stimuli shaped by contact location, feature type, and active touch speed. Our analyses show that cockroach-inspired antenna mechanics jointly with active touch speeds improve feature classification accuracy compared to conventional sensors with uniform flexural stiffness gradient by increasing tactile data sparsity and dispersion. An exploration of stiffness and damping of antenna mechanics revealed design trade-offs that influence tactile discrimination and structural stability. Through sim-to-real transfer, stiffness gradients and structured active touch motions were demonstrated on a miniature distributed soft robotic antenna, validating their effectiveness in real-world robotic systems. Taken together, this work presents a biologically grounded framework for tactile sensor design that reduces computational load and enhances adaptability.