A Cyclic Negative Pressure Therapeutic Robotic Approach With Contact Pose Correction and Adaptive Mouthpart Variation for Limbs Suction

Abstract

Negative Pressure (NP) therapy has become a widely adopted treatment for extremity diseases, with growing demand due to increasing prevalence of limb disorders. To address healthcare workforce shortages, this paper presents a robotic system for automated cyclic NP therapy capable of handling morphologically varied limbs and operational challenges including sensor noise, control inaccuracies, calibration errors, and unpredictable patient motion. The system integrates a vacuum therapy head with haptic feedback and adaptive mouthpart geometry from prior work, enhanced with a novel hybrid force/impedance controller for stable limb interaction. Visual detection generates initial suction poses constrained by the mouthpart dimensions, while real-time tactile feedback enables pose correction during contact. Through phantom and human experiments, this system has been validated with the efficacy of NP suction and is robust to system errors and people’s movement. In the pose correction experiment, the robotic system modified the contact pose with different orientations through the contact force control to make the suctions on the standard phantom. Then, compared to the other methods, the proposed method was utilized in the phantom and human arm suction experiments to apply suctions on unstructured arms with a higher success rate and achieve a higher comfort level.