Within the realm of intelligent manufacturing, industrial robots occupy a paramount position as electromechanical keystones. Consequently, the scrutiny of advanced state monitoring methodologies for these robotic systems is of utmost significance. Currently, dynamic analysis based on vibration signal has been used to evaluate robot macroscopic behavior. Nonetheless, these tools fall short for the rigorous scrutiny demanded by such endeavours, which necessitate a meticulous dissection of each subcomponent’s functionality. This paper aims to fill this gap by combine traditional dynamic analysis methods with robot joint state monitoring, containing an accurate description of its full dynamic characteristics. Harmonic reducer was used to transfer torque, so that arm can move. To conclude, the real robot on a test rig were tested with a set of different condition and speed. The experimental results highlight the ability of the dynamic analysis method to accurately replicate joint angular rotation and speed in a wide range of operational scenarios. The amplitude of characteristic frequency can be used to monitor the running state of robot joint. This research endeavours to construct a comprehensive maintenance CM-oriented framework, tailored for conducting in-depth and sophisticated analyses that unravel the intricate impacts of manipulator degradation.