Multi-modal activity-based mobility generation applied to microscopic traffic simulations

In recent years, we collectively decided that to achieve sustainable growth in our cities, we are going in the direction of smart-city and smart-mobility studies. With the increasing amount of data available and the presence of extensive communication infrastructure for vehicles and people to connect, many efforts are made towards mobility optimizations. Applications such as personal trip planners and similar are in widespread use at the moment, and they have an impact on mobility patterns in and around a city. Using a mobility simulator framework it is possible to study their impact on the overall traffic congestion, and possibly develop an intelligent and coordinated trip planner that keeps into account different modes of transportation and parking availability that would satisfy the user without impairing the system. To achieve a deployable solution, we need to base our optimizations on realistic topology representations and realistic mobility models of our modern cities. Although significant progress has been made on the realism of mobility simulators, substantial work needs to be done to achieve realistic mobility patterns. We aim to build a comprehensive framework that allows the creation of a realistic topology and infrastructure for a city and to generate realistic mobility for it. We selected SUMO1 microscopic mobility simulator because it is open-source, flexible and performant, with the advanced multi-modal featured that we require, and most importantly, the user and developer community is very active and always ready to help. The first part of this framework is already available on GitHub under GPLv2 license, and it comprises of (i) the toolchain used to generate a scenario starting from OpenStreetMap2 data, (ii)

the toolchain used to generate reasonable mobility based on a representation of origin-destination time-dependent matrices3, and (iii) our case study of the Principality of Monaco and its surroundings4. The Monaco SUMO Traffic Scenario is available at https://github.com/lcodeca/MoSTScenario. Despite that the mobility generator already implemented provides reasonable mobility features, it is not detailed enough when it comes to personal trip planning, activity-based mobility, and parking patterns and behaviors. The new tool we are implementing now is meant to tackle these challenges.