Abstract: To address the operational problems including frequent geological disasters, difficult emergency rescue, low passenger flow and high operating costs of mountain tourism rail transit, this paper proposed and built an operation support platform for mountain tourism rail transit integrating functions such as disaster perception, emergency response and intelligent maintenance. It relied on microservice architecture and Internet of Things (IoT) technologies to construct a real-time monitoring system for multi-source data covering meteorology, geology, tracks, vehicles and other fields. This study fused Kafka streaming data via risk assessment models, realized disaster risk early warning and train operation auxiliary decision-making, and meanwhile automatically generated closed-loop maintenance tasks. Experimental results indicated that the platform could accurately identify typical safety risks such as heavy rainfall and slope displacement in the test section of Dujiangyan-Siguniangshan mountain rail transit, conducted data analysis and issued scheduling suggestions. It effectively improves emergency disposal efficiency and operation and maintenance collaboration level, achieves the intelligent operation and maintenance goals of lean staffing, high efficiency and low cost, and provides crucial technical support for the safe and long-term operation of mountain rack railways.