Soft Robotic Proprioception Enhancement Via 3D-Printed Differential Optical Waveguide Design

Shaowu Tang,Kailuan Tang,Yutao Guo,Shijian Wu,Jiahao Xu,Benkang Lou,Sicong Liu,Juan Yi,Jian S. Dai,Zheng Wang

DOI: https://doi.org/10.1109/lra.2024.3416795

IF: 5.2

2024-01-01

IEEE Robotics and Automation Letters

Abstract:Soft robots undergo complex deformations during actuation and interaction due to the flexibility and compliance of their soft materials. This characteristic presents challenges in proprioception, particularly in characterizing their spatial deformations. Soft optical waveguide sensors have emerged as a reliable solution for soft robotic proprioception owing to their compact form factor that enables ease of integration into the soft body, and their high sensitivity in detecting deformations. In this study, we introduce a differential optical waveguide design that combines light-blocking and light-transmitting structures featured with distinct light attenuation, leading to monotonically variable sensing signals in response to bidirectional deformation. Consequently, this design enables the differentiation of two directions of deformation using merely one sensor. We employ three-dimensional printing technology and adjust printing parameters to tune the attenuation of both structures and accommodate various deformation modes, including proprioceptive bidirectional bending, shearing, and twisting. Furthermore, the sensing capabilities can also be extended to cope with deformation in three dimensions, based on the combinations of differential optical waveguides. The proposed method is validated through design, simulations, fabrication, and experiments. Finally, a differential optical waveguide is integrated into a dual-chamber soft actuator showcasing the performance in detecting the bidirectional bending motion. This developed methodology offers design guidance for soft optical sensors and may inspire the development of integration of sensing in soft robots capable of complex motion.