Safe Contact Control of Acupuncture Robot Based on Vision and Admittance Control

Abstract

Traditional Chinese Medicine (TCM), particularly acupuncture treatments, enjoys widespread popularity in China and has been effective in addressing various symptoms. However, the practice of TCM requires a high level of expertise, and performing these procedures manually can result in clinician fatigue. Moreover, the uneven distribution of medical resources and a shortage of qualified medical professionals across China, coupled with the variability in technical proficiency among practitioners when performing these procedures, present significant challenges. To address these issues, the development of acupuncture robots presents an innovative solution, aiming to automate the process of acupuncture. However, the main challenges faced by acupuncture robot technology include ensuring safe and gentle interaction between the needle and skin, as well as avoiding safety issues caused by minor patient movements. To address these challenges, this study developed a platform integrating mini cameras, a six-degree-of-freedom collaborative robotic arm, and a six-dimensional force sensor, combined with our proposed YOLOv8n-Ghost keypoint detection algorithm and admittance controller, to achieve safe contact control of the acupuncture robot after needle insertion. The results of this study validate the effectiveness of the control scheme, providing an important reference for enhancing the safety and reliability of acupuncture robots.