Abstract: This paper systematically investigates robot adaptive insertion and measurement technology for complex and precise plugging scenarios, aiming to overcome the limitations of traditional rigid automation solutions and achieve compliant and intelligent evolution in manufacturing processes. Focusing on "compliant insertion," it thoroughly explores key technologies across the full chain from mechanism design to perception-cognition and control-decision-making. A multimodal perception system integrating vision and force-position control is constructed to enable accurate estimation of contact states and relative pose errors. Through the establishment of an intelligent robotic industrial experimental platform and batch deployment applications, the proposed theories and methods are validated. The developed technology significantly enhances the success rate, efficiency, and safety of robotic plugging assembly tasks in high-precision environments, providing new technical pathways and theoretical foundations for the evolution of intelligent manufacturing.