Soft Robotic Grasping Perception System Based on Multimodal Intelligent Sensor Fusion

Abstract

. Soft robots have become a key technology for precise grasping and human-robot collaboration in complex environments due to their flexible structures and high safety. Multimodal intelligent sensor fusion has shown significant advantages in perception and decision-making of soft robots. Soft robots can achieve high-precision grasping and optimized movements in dynamic, obstructed or uncertain environments by integrating tactile, visual and inertial data from multiple sources and combining it with multimodal intelligent sensing algorithms. At the hardware level, tactile, visual and inertial sensors form the basis of multidimensional perception. At the algorithmic level, fusion algorithms, adaptive fusion strategies, and learning-driven action optimization have enhanced the robustness and adaptability of the system. Multimodal intelligent sensor fusion technology is widely used in scenarios such as industrial precision operations, rehabilitation assistance, and underwater grasping. This article explores the development trends and future research directions of multimodal intelligent integration in soft robots. It provides a reference for the next generation of highly autonomous and adaptable intelligent soft robot systems.