AI-Driven Energy Optimization for Mixed Transport Systems: Enhancing Sustainability and Cost Efficiency in Urban Mobility

The escalating demand for urban transportation amid rapid urbanization requires sustainable solutions that optimize energy consumption across multiple transportation modes. Although recent studies have applied Artificial Intelligence (AI) to public transportation, electric vehicles (EVs), or Mobility-as-a-Service (MaaS) individually, limited research has addressed the integrated optimization of mixed transport systems. This article proposes the Mixed Transport Energy Optimizer (MTEO), an AI-IoT framework integrating Multi-Agent Reinforcement Learning (MARL), Graph Neural Networks (GNNs), Long Short-Term Memory (LSTM) networks, IoT-based sensing, and edge computing for real-time optimization of routing, scheduling, and energy management across high-capacity public transport and low-capacity on-demand mobility. The framework dynamically adapts to passenger demand, traffic conditions, energy pricing, and environmental factors, while leveraging V2G-enabled energy management and predictive demand forecasting. MTEO is evaluated using the Mumbai mixed transportation network through SUMO and MATLAB/Simulink simulations, supported by a preliminary corridor-level pilot validation. The proposed framework achieves 23 percent reduction in daily energy consumption (4.2–3.24 GWh), 26 percent reduction in operational cost, 25 percent reduction in CO₂ emissions, and 28 percent reduction in passenger delay compared with conventional transportation systems. These improvements demonstrate the effectiveness of coordinated cross-modal optimization for sustainable urban mobility. The proposed MTEO framework provides a scalable solution for intelligent transportation systems (ITS), supporting Sustainable Development Goals (SDGs) and advancing the transition toward net-zero smart cities.