HFR-Video-Based Vibration Analysis of a Multi-Jointed Robot Manipulator

Abstract

As the demand for industrial robots continues to increase, monitoring robot manipulators in factory environments has become essential to ensure proper and precise operation. Unexpected vibrations can reduce the production efficiency and quality, causing financial losses, and safety risks to workers. Evaluating a robot’s vibration resistance solely through arm movements makes it challenging to accurately capture fine vibration-frequency responses using conventional methods. Traditional analyses rely on contact sensors, which are limited by the number of measurable points, and often involve high costs. In this study, we employed high-frame-rate (HFR) cameras for non-contact vibration analysis, enabling a detailed evaluation of the vibration characteristics during robot operation. By processing the 500 fps HFR video using digital image correlation, we analyzed the frequency responses of sub-pixel displacements at multiple locations and quantified changes in the vibration amplitude and phase across different parts of the robot. This approach provides a more precise understanding of fine vibration distributions and their impacts. The proposed method is accurate and can simultaneously measure multiple points.