Search papers, labs, and topics across Lattice.
This paper introduces a 3D-printed soft optical sensor (SOLen) that uses a printed lens to guide light within a Y-shaped waveguide for mechanosensing. The sensor's mechanism relies on deformation-induced lens rotation and focal-spot translation, which redistributes optical power between the waveguide branches, enabling the encoding of both motion direction and amplitude. The authors demonstrate reproducible branch-selective signal switching through rotational tests, establishing a workflow for creating soft optical sensors with integrated lenses.
A 3D-printed lens can focus light within a soft robot's optical sensor, enabling directional mechanosensing without multi-material interfaces.
Additive manufacturing is enabling soft robots with increasingly complex geometries, creating a demand for sensing solutions that remain compatible with single-material, one-step fabrication. Optical soft sensors are attractive for monolithic printing, but their performance is often degraded by uncontrolled light propagation (ambient coupling, leakage, scattering), while common miti- gation strategies typically require multimaterial interfaces. Here, we present an approach for 3D printed soft optical sensing (SOLen), in which a printed lens is placed in front of an emitter within a Y-shaped waveguide. The sensing mechanism relies on deformation-induced lens rotation and focal-spot translation, redistributing optical power between the two branches to generate a differential output that encodes both motion direction and amplitude. An acrylate polyurethane resin was modified with lauryl acrylate to improve compliance and optical transmittance, and single-layer optical characterization was used to derive wavelength-dependent refractive index and transmittance while minimizing DLP layer-related artifacts. The measured refractive index was used in simulations to design a lens profile for a target focal distance, which was then printed with sub-millimeter fidelity. Rotational tests demonstrated reproducible branch-selective signal switching over multiple cycles. These results establish a transferable material-to-optics workflow for soft optical sensors with lens with new functionalities for next-generation soft robots