A Bioinspired Multimodal Soft Gripper With Intuitive Teleoperation and Distal Sensing for Constrained-Space Manipulation

Abstract

Inspired by the multimodal deformation and distal sensing capabilities of the hawkmoth proboscis, this work presents a bioinspired multimodal soft gripper for intuitive teleoperation and distal sensing for constrained-space manipulation. The gripper integrates dual fabric-based pneumatic chambers with a rigid helical spring skeleton, enabling a compact coiled state at rest and controllable unfolding, straightening, and coupled wrapping during operation. The proposed design achieves a maximum reachable depth of 210 mm and a maximum pull-off force of 7.5 N, providing both large-scale deployment and sufficient load-bearing capability for narrow and difficult-to-access environments. A finger-worn interface is developed to control the two pneumatic chambers independently or synergistically through a stable mapping between finger flexion and injected air volume. The shape-control accuracy and repeatability are further evaluated over commanded finger angles from 15° to 90°. To enhance safety and robustness, a distal pressure-sensing unit enables real-time contact detection and triggers a finger-worn pneumatic airbag for haptic feedback, forming a human-in-the-loop grasping strategy. Compared with open-loop operation, the haptic-feedback mode reduces deformation of compliant objects, increasing its post-grasp diameter from 17.75 to 22.11 mm relative to an initial diameter of 25.03 mm. The complete system is validated through confined-space grasping and pick-and-place task, close-contact human–robot interaction, as well as quadruped-mounted outdoor branch-clearing and object-retrieval tasks. Experimental results demonstrate that the proposed gripper combines intrinsic compliance, multimodal kinematic adaptability, distal perception, and intuitive human-in-the-loop control, highlighting its potential for service robotics, human–robot collaboration, and field operations in cluttered and constrained environments.