Efficient Path Planning of Soft Robotic Arms in the Presence of Obstacles
Preston Fairchild, Vaibhav Srivastava, Xiaobo Tan
Abstract
Soft robotic manipulators have seen growing interest in recent years and have many applications in the medical and industrial fields. Path planning algorithms for these soft continuum arms often have features to include obstacle avoidance. The redundant nature of soft robots allows for these manipulators to avoid obstacles while moving towards a goal. In this paper, a novel, efficient path planning algorithm is proposed for a soft robotic arm to navigate multiple obstacles in its workspace. The method aims to reduce the computational complexity and increase the precision of modeling curve-like obstacles by representing them with parametric equations instead of a set of points. The closest point between a function modeling the robotic arm and a function approximating the obstacle is updated in real-time and used to accommodate obstacle avoidance in path planning. The method is tested in simulation with a soft continuum arm represented by the piecewise-constant curvature model and the performance is compared to the traditional approach where an obstacle is defined by a set of points. The efficacy of the proposed approach is supported by simulation results from multiple obstacle settings, including one emulating multiple tree branches.