Robot Motion Planning and Control for Thread Hole Cleaning Tasks

Abstract

Painting is a necessary process for metal surface protection. However, the painting material may accumulate in the thread holes and cause difficulties for future bolt installation. Cleaning the extra paint is necessary for the following assembly task. For industrial robots, cleaning such paint in internally threaded holes is not a trivial task. It requires overcoming challenges related to the accuracy of hole positioning and deviations in the orientation of the threaded holes during the cleaning process, otherwise, the tools may be jammed or the threads may be damaged. This paper aims to analyze the position and force constraints in this process and explore the method to enable robotic automatic thread cleaning. The whole cleaning task is divided into 5 processes: approaching, hole searching, posture tuning, thread cleaning, and retraction. Two spiral search methods are proposed to compensate for the translational and rotational error between the tab and the threaded hole; A force control is introduced to synchronize the rotation and feeding motion; A switched hybrid force/positional controller is proposed to accomplish the cleaning task, which ensures the motion accuracy and compliance. The proposed method has been implemented on an experimental platform, and the results verify the effectiveness of the proposed threaded hole cleaning method.