Rapid growth in the demand for urban transportation networks has underlined the urgent need for sophisticated management of multimodal transportation networks due to the significant costs associated with traffic congestion and CO2 emissions. Addressing these challenges requires a dual focus: alleviating congestion and minimizing CO2 emissions. A significant step towards achieving environmentally sustainable urban transportation networks, SMART-ROUTE is dedicated to optimizing large-scale urban multimodal transportation networks and has four objectives: (i) capturing the evolution of large-scale multimodal network dynamics; (ii) design and train machine learning models to predict the dynamic network state; (iii) design Deep Reinforcement Learning (DRL) methods to refine demand and supply management; (iv) validate the proposed algorithms at large-scale multimodal transport networks. Calculation of the real transportation network state is defined as an equilibrium that corresponds to modeling and solving Dynamic network equilibrium (DNE). It addresses the challenges posed by DNE, with increasingly fluctuating demand and supply, particularly in light of the health crisis and the spread of teleworking. Utilizing machine learning approaches, such as physics-informed machine learning (PIML) and graph neural networks (GNNs), we aim to capture intricate DNE features (e.g., path/link flow patterns), enabling more efficient predictive modeling. Graph-based machine learning models will be designed for forecasting DNE, while DRL techniques optimize these solution spaces by applying various control strategies. SMART-ROUT will evaluate and validate the methods on multiple multimodal test cases. In particular, real test cases of Paris and Lyon, provided by Consortium members, will be employed, ensuring their robustness and efficacy. By integrating the DNE mathematical foundations with the power of ML, SMART-ROUTE aims to create new efficient tools to optimize agent-based simulators. This integration leads to more adaptive, efficient, and environmentally conscious urban transport networks that are responsive to a variety of conditions, advancing the transition to smart, sustainable cities. The tool will offer time, energy, cost, and CO2 emission savings for professionals in the transportation and urban mobility sectors. The project will offer a set of models and tools available to all researchers, policymakers, and transportation stakeholders.