"As cities grow rapidly and more people through different modes compete for limited urban road infrastructure to travel, it is important to manage traffic space to improve accessibility for travelers. This project tackles the problem of modeling and optimization in large-scale congested traffic networks with an aggregated realistic representation of dynamics and route choice and multiple modes of transport. This is a highly motivating problem both because of the socio-economic influence of congestion and the challenges embedded in the optimization framework and the modeling aspects. Currently most optimization methods for transport networks (i) are suited for toy networks with simplified dynamics that are far from real-sized networks, (ii) apply decentralized control, which is not appropriate for heterogeneously loaded networks, (iii) investigate engineering solutions through micro-simulation models and scenario analysis that make the problem intractable in real time, (iv) are not considering interactions and conflicts between transport modes (car, bus, delivery vehicle). This problem is even more challenging if one considers that transportation networks have a hierarchical structure with freeways and urban roads with mixed or separated traffic (e.g. bus-only lanes), that have dissimilar traffic flow dynamics. Lack of coordination among the jurisdictions during traffic operations or limited means of traffic data monitoring and communication can impede such mixed traffic network ideal goal. Traditionally, choices of people in transportation networks are based on equilibrium conditions with small variations.The huge amount of datasets (including thousands of GPS data from taxis, cars and buses and road detector data from heavily populated cities worldwide) can provide a unique way to understand how really people make choices, how these choices affect the development and spreading of congestion in networks and integrate them in the macroscopic dynamics and optimization"