Weighted path planning based on collision detection

Abstract

Purpose Aims to make an industrial robot work human‐friendly while coexisting with humans in a same working space, without any modification of the manipulator hardware as well as its motion controller. Design/methodology/approach Presents a weighted path planning approach based on collision detection for a robot sharing the same workspace with humans. Using a base and wrist force/torque sensors, a model‐based method is investigated to estimate the collision force and collision position on the manipulator; then a weighted path planning approach is developed to control the human‐robot interaction. Simulation experiments, with collisions between the manipulator and static objects and moving objects, are conducted to validate the efficacy of the method. Findings The experimental results illustrate the validity of the developed collision detection and planning scheme and make it possible for industrial robots to work safely around humans. The proposed weighted path planner (WPP) outperforms other three path planners. Originality/value Introduces the WPP based on collision detection. The wrist and base force/torque sensors configuration has the friction free benefit, and the developed method does not modify the existing designs of industrial robots. The contact force is well controlled under the human pain tolerance limit, through the modification of the desired path, and does not need torque control which is usually not available to industrial robots.