By combining advanced modeling, data science and systems optimization, the program prepares students to become pioneers in transportation research and innovation, contributing to sustainable, resilient and intelligent mobility systems.

This course focuses on the principles of macroscopic traffic flow dynamics theory, providing students with a comprehensive understanding of large-scale transport systems. Key topics include link dynamics based on the Lighthill-Whitham-Richards (LWR) model, which captures the evolution of traffic flow on road links. It also explores local dynamics at network level using node modeling to analyze traffic behavior at intersections. In addition, the course covers static and dynamic traffic assignment techniques to model and optimize vehicle distribution on a network. Finally, it looks at overall network dynamics using the Macroscopic Fundamental Diagram (MFD) to give an overview of the overall performance of urban transport systems.

The course “Micro-mobility Traffic” focuses on the growing impact of micro-mobility vehicles (MMVs), such as e-bikes and e-scooters, in urban areas. These vehicles offer promising solutions to reduce car usage and its associated externalities but also bring new challenges, especially in their interactions with pedestrians in shared spaces. This course aims to provide students with a deep understanding of MMV rider and pedestrian behavior, develop mathematical models to describe their interactions, and assess the implications for traffic safety and congestion. Key learning outcomes include defining and analyzing behavior characteristics, designing experiments and data collection methods for MMVs and pedestrians, modeling movement and behavior, calibrating and validating simulation models, and designing infrastructure tailored to MMVs and pedestrian needs. By the end of the course, students will be equipped to evaluate and plan for future scenarios with increased MMV usage.

This course focuses on dynamic traffic simulation tools, which are essential for understanding the functioning of road networks and diagnosing potential inefficiencies. These tools are also widely used to assess the impacts of new infrastructure developments or dynamic regulation strategies. The course provides an overview of dynamic simulation tools, with a particular emphasis on positioning the MnMs platform [MOU1] and its use cases. Students will learn how to run simulations, analyze the results, and develop mobility management scenarios. Additionally, the course highlights how to use simulation as a decision-making tool and equips students with the skills to interpret and effectively communicate the outcomes of their analyses.

The MLP course aims to introduce the process of data analysis, processing, and interpretation, as well as the application of machine learning techniques. The course presents a broad overview of different supervised (classification and regression) and unsupervised (clustering, data reduction) learning methods. The various techniques are used to understand the hidden structures and regularities (patterns) present in data and to develop classification or prediction models taking advantage of large datasets from different application domains (transport, socio-economics, geography, etc.).

Several datasets will be made available to students and processed in Python, using several available open-source libraries for data processing and machine learning (e.g. numpy, pandas, matplotlib, scikit-learn, networkx, etc.). The course adopts a hands-on approach, with a project (Python) that will be developed throughout the course.

The DL-DNA course offers a comprehensive introduction to the principles and applications of modern deep learning. It begins with the fundamentals of artificial neural networks and progresses to explore advanced architectures and learning paradigms. A significant focus is placed on graph representation learning, covering machine learning on graphs, graph learning tasks, and graph neural network architectures. The course also delves into real-world applications, showcasing how graph neural networks and advanced deep learning techniques are utilized in science and engineering. By the end of the course, students will have a strong theoretical and practical understanding of cutting-edge deep learning methodologies and their application to complex problems in various fields.

The course MCB explores the critical role of understanding individual behavior in predicting demand and anticipating changes driven by the introduction of new services and products. Over the past decades, discrete choice models have become a cornerstone in various fields, including transportation, marketing, environmental economics, and health. This course provides students with the theoretical foundations of choice models and hands-on experience applying these models to engineering case studies. Key topics include data collection methods, behavioral theories, and the mathematical principles underlying choice modeling. Students will develop logit models based on alternative attributes and socio-economic factors, estimate these models using Python software, and assess the statistical significance of variables. By the end of the course, participants will be able to calculate forecasting indicators and predict individual responses to future scenarios, leveraging choice models as powerful tools for decision-making.

This course introduces the fundamental and advanced concepts of computer vision. It enables the processing, analysis and interpretation of images and videos using convolutional neural network (CNN) algorithms for object recognition. This will be followed by an in-depth look at object detection with YOLO, a fast and efficient approach to real-time detection. There will also be a focus on image calibration and colorimetry techniques, as well as lighting system design, to optimize image acquisition and processing for better performance of computer vision models. A “fil rouge” project will apply these concepts to the detection and recognition of objects such as traffic signs.

The Mobility Data Hackathon is an intensive, collaborative event where students can apply the knowledge and skills acquired in the “data science for mobility” and “mobility modelling” UE’s. Over two days, participants will work in teams to solve real-life mobility problems using cutting-edge data science techniques. Tasks may include analyzing mobility patterns, developing predictive models for traveler behavior, or designing innovative solutions for sustainable and efficient urban transportation. Drawing on tools such as deep learning for image and data analysis, choice modeling to understand user preferences, machine learning for advanced predictions, and MnMs simulation platform, students will tackle complex problems and present their solutions to a panel of experts. This hands-on experience offers a unique opportunity to showcase creativity, teamwork and technical expertise while tackling pressing mobility and transportation issues.

The “Mobility Control and Management – MoCoM” course is designed to introduce students to some of the most common procedures for controlling and managing people’s mobility, leading to the concept of Intelligent Transport Systems (ITS). The course focuses on different management scales (intersections, neighbourhoods, etc.) and environments (urban, highway). The course begins with a general introduction to the historical issues and their evolution with the emergence of Information and Communication Technologies (ICT), leading to the development of Intelligent (and Cooperative) Transport Systems. These now offer new opportunities to think, model and regulate mobility in a multi-modal, sustainable, resilient and cooperative way. An overview of the different ways of thinking about and managing mobility is then offered, depending on the environment (urban, motorway), scale (intersection or urban area) and/or type of manager, with a focus on the techniques employed in relation to the problems addressed. Finally, a panel of the main families of commonly used tools (control loops, predictive models and reinforcement learning) and classically implemented tools is presented. Students are then invited to validate their knowledge by putting it into practice through exercises on Python. In the final session, students defend their results by presenting them to their colleagues.

The course DOTNET provides students with a comprehensive understanding of the principles and methods for planning efficient, multimodal transport networks. The curriculum covers key topics such as transportation network and demand representations, network loading, and equilibrium analysis, enabling students to model and evaluate the performance of complex systems. Students will explore multimodal transport network design, accessibility analysis, and strategies to enhance network resilience. The course also delves into future trends shaping transportation, including electrification, automation, and shared mobility. Through case studies and practical applications, participants will gain hands-on experience in designing sustainable and resilient transportation systems tailored to evolving urban needs.

The course focuses on the assessment of externalities associated with transportation systems, including noise pollution, environmental pollution and contributions to climate change through greenhouse gas emissions. Students will learn to quantify and analyze the impacts of mobility on urban environments and the wider ecosystem. The course covers methods for assessing noise levels, calculating pollutant emissions and estimating transport-related greenhouse gas emissions. It also explores strategies for mitigating these externalities, promoting sustainable transport solutions and supporting climate action objectives. By combining theoretical foundations with practical tools, this course enables students to critically assess and address the environmental challenges posed by modern mobility systems.