A predictive maintenance model for sustainable urban transit: the Riyadh Metro case study

The Riyadh Metro is a prominent initiative in Saudi Arabia, aimed at revolutionizing urban transportation and alleviating traffic congestion in the city. Comprising six metro lines and 85 stations, the network is projected to accommodate millions of passengers each day, which requires innovative maintenance strategies to ensure reliable and efficient operation. This paper proposes a real-time health monitoring framework that integrates Internet of Things (IoT) sensors, digital twin technology, and a Temporal Graph Neural Network (TempGNN) model to predict and prevent equipment failures. IoT sensors installed on critical train components, e.g., brakes, wheels, engines, Heating, Ventilation, and Air Conditioning (HVAC) systems, continuously feed data to a digital twin, a live virtual replica of the metro system, enabling continuous condition tracking. The TempGNN exploits both temporal patterns and spatial relationships in these sensor data to forecast potential failures with high accuracy ≈ 98%, allowing maintenance to be scheduled proactively. In testing, the proposed system reduced unplanned downtime by over 30% and maintenance costs by about 25%, while improving energy efficiency and passenger satisfaction. By anticipating faults and optimizing maintenance schedules, the framework enhances the metro’s reliability and safety. The approach demonstrates considerable benefits for large-scale transport networks and can serve as a model for implementing predictive maintenance in metro systems worldwide.