With the continuous expansion of the scale of urban rail operation, the construction of intelligent operation and maintenance system based on condition monitoring, feature extraction, condition assessment, fault diagnosis, fault early warning, optimization of maintenance methods and decision-making fusion is the development trend of urban rail transportation operation and maintenance.

Since the inauguration of the on-board PIS system of Suzhou Rail Transit Line 2 in 2013, although it has realized partial data monitoring and storage, it is unable to achieve real-time monitoring and remote control of equipment status, and lacks data mining analysis. The maintenance method is more traditional, and no early warning can be made for potential equipment failure.

In order to solve the problems of the operation and maintenance of the traditional on-board PIS system, such as insufficient real-time monitoring of faults, most of which are found after the event, untimely emergency disposal, and no hidden danger mining and troubleshooting mechanism, the research project of Artificial Intelligence for IT Operations was carried out in Suzhou Metro Line 2, and a small batch trial installation application was formed. The operation status, network status, system performance, data storage status, etc. of the on-board PIS system equipment were monitored in real time, Implement on-site replaceable components for fault detection and positioning, failure warning for key components, and remote control.

The main functions of the system are data acquisition, data processing, real-time monitoring, fault diagnosis, fault early warning, statistical analysis, health assessment, maintenance management, equipment management, decision-making management and remote control.

The expert group of Shanghai, Suzhou and other rail transit industries in East China carefully reviewed the acceptance materials and listened to the report of the project team, and agreed that the acceptance materials are complete and comprehensive, the research method is reasonable, and it meets the acceptance criteria of the project, and the expert group unanimously agreed that the project passes the acceptance evaluation.