Active collision avoidance for teleoperated multi-segment continuum robots toward minimally invasive surgery

Abstract

Collision avoidance presents a challenging problem for multi-segment continuum robots owing to their flexible structure, limited workspaces, and restricted visual feedback, particularly when they are used in teleoperated minimally invasive surgery. This study proposes a comprehensive control framework that allows these continuum robots to automatically avoid collision and self-collision without interfering with the surgeon’s control of the end effector’s movement. The framework implements the early detection of collisions and active avoidance strategies by expressing the body geometry of the multi-segment continuum robot and the differential kinematics of any cross-section using screw theory. With the robot’s parameterized shape and selected checkpoints on the obstacle’s surface, we can determine the minimum distance between the robot and arbitrary obstacle, and locate the nearest point on the robot. Furthermore, we expand the null-space-based control method to accommodate redundant, non-redundant, and multiple continuum robots. An assessment of the avoidance capability is provided through an instantaneous and global criterion based on ellipsoids and possible movement ranges. Simulations and physical experiments involving continuum robots of different degrees of freedom performing various tasks were conducted to thoroughly validate the proposed framework. The results demonstrated its feasibility and effectiveness in minimizing the risk of collisions while maintaining the surgeon’s control over the end effector.