［Objective］Aiming to design and construct a high-performance, high-reliability intelligent O&M (operation and maintenance) system for urban rail transit communication and to formulate scientific maintenance strategies, it is necessary to transform traditional periodic scheduled maintenance to condition-based maintenance according to test analysis, and to prior assessment-based predictive maintenance, thereby reducing the probability of equipment failure and operational costs. Furthermore, it is to shift from traditional manual O&M management to an automated, information-based intelligent monitoring and maintenance approach.［Method］In combination with the on-site practice, associations are established between virtual entities and physical entities such as communication systems and key equipment to investigate equipment monitoring and diagnosis models. The multi-source heterogeneous data fusion and analysis technologies are applied to communication system fault diagnosis and prediction. An equipment health index algorithm model and a big data architecture are proposed, and a tracking and evaluation platform encompassing ′cloud-data-intelligence′ is built. This platform provides both a theoretical basis and a validation environment for on-site pre-diagnostic maintenance, enhances the coverage of condition awareness, the accuracy of diagnosis and prediction, thus significantly improving the quality of O&M.［Result & Conclusion］The engineering application of the intelligent communication intelligent O&M system on Nanning Rail Transit Line 4 achieves comprehensive equipment monitoring, extensive fault response mechanism, and advanced human-machine interaction. Application results show that this intelligent communication O&M system enhances the intelligent O&M level of rail transit communication equipment and enables a shift from traditional periodic maintenance to health-based intelligent monitoring and predictive maintenance.