Modeling Vehicular Flow with Traffic Light Cycles in Quito

Abstract

We introduce a conceptually simple agent-based model to investigate vehicular traffic in the city of Quito, Ecuador, with a particular focus on how traffic light cycle configurations influence average travel times. The model captures essential features of urban traffic dynamics through an agent-based framework and has been efficiently implemented in Python. Our results show that traffic light cycles play a decisive role in shaping travel times: while certain configurations substantially reduce delays for most agents, others may increase travel times for some agents, especially those covering shorter routes. By explicitly accounting for interactions between multiple traffic lights, we identify configurations that minimize the average travel time across the network. We also find that it is possible to achieve relatively uniform travel times among agents, a crucial aspect for improving overall traffic performance without disproportionately affecting specific user groups. Although the model is implemented using the street network and traffic light system of Quito, its structure is general and can be readily applied to urban traffic scenarios in other cities with limited resources.