Abstract: This paper systematically reviews the latest application advances of intelligent inspection robots in the field of rail transit vehicle maintenance. The research focuses on the system architecture, core technologies, and application effectiveness of intelligent inspection robots. It provides an in-depth analysis of their technical principles and implementation pathways in key areas such as identification and localization of undercarriage components (based on hierarchical neural networks and deep learning), precise localization of abnormal states, and 3D dimensional measurement (combining binocular stereo vision and Fourier Profilometry). The article highlights the significant achievements of intelligent inspection robots in optimizing maintenance processes, extending inspection cycles, and improving maintenance quality and efficiency. It further discusses current challenges and future development directions. Practical evidence demonstrates that the application of intelligent inspection robots delivers substantial socio-economic value and serves as a pivotal technology for advancing the intelligent transformation of rail transit operation and maintenance.