On September 6th, 2022, Daniele Gammelli successfully defended his Ph.D. thesis entitled “Learning and Control for Adaptive Transportation Systems.”

Congratulations, Daniele!

He was supervised by Professors Francisco Camara Pereira, Dario Pacino, and Filipe Rodrigues.

Examiners were Profs. Evelien van der Hurk (DTU), Alexandre Alahi (EPFL), and Guido Gentile (Sapienza) .

The defense session was chaired by Prof. Stefan Eriksen Mabit.

Daniele published 6 interesting papers from his Ph.D. thesis! You can find below the abstract of the thesis and the list of papers. For more information, feel free to contact Daniele at daniele.gammelli@gmail.com or directly via his Linkedin page.

Abstract: Transportation is a permeating factor of modern society, effectively enabling the accomplishment of everyday tasks such as going to work, school, or simple personal errands. The efficiency of the transportation network highly influences the quality of life of individuals, and it is widely recognized that investments in transport infrastructure can generate large developmental payoffs throughout society.
Currently, transportation within cities is intrinsically system-first, whereby users have to adapt to a fixed transportation system (e.g., fixed bus routes and schedules). In contrast, we believe in the vision of a user-first system, whereby supply and demand for transportation are seamlessly co-adapted into an Adaptive Transportation System (ATS). To pursue this vision, transportation systems need to acquire two fundamental capabilities: 1) being able to understand and predict current mobility patterns, and 2) converting this knowledge into effective decisions that can satisfy the need for mobility of a broad and diverse audience. At the same time, the high complexity of the transportation system, together with the vast availability of mobility data (e.g., consider the daily stream of data generated by our smartphones), make this a perfect scenario for applications of machine learning and data-driven strategies broadly. In this context, we aim to develop learning algorithms that are capable of addressing the challenges emerging within adaptive transportation systems, with a focus on mobility-on-demand systems.
In the first part of the thesis (i.e., “Perception and Prediction”) we develop tools for predicting the evolution of mobility demand. In the second part of the thesis (i.e., “Optimization and Control”) we develop autonomous decision-making strategies for an efficient allocation of supply. To make such systems a reality, autonomous service operators must be able to learn about their environment via sensing and interaction and use the results of this learning to improve performance or enable safe operation.
In other words, the goal to expand the capabilities of autonomous transportation systems means that the next generation of these systems is fundamentally a generation of learning operators. This thesis attempts to bridge the gap between current and future transportation systems.

List of Papers:

• Paper 1: Estimating latent demand of shared mobility through censored Gaussian Processes (https://doi.org/10.1016/j.trc.2020.102775)
• Authors: Daniele Gammelli, Inon Peled, Filipe Rodrigues, Dario Pacino, Haci A. Kurtaran, Francisco C. Pereira
• Journal: Transportation Research Part C: Emerging Technologies

• Paper 2: Recurrent Flow Networks: A Recurrent Latent Variable Model for Density Estimation of Urban Mobility (https://doi.org/10.1016/j.patcog.2022.108752)
• Authors: Daniele Gammelli, Filipe Rodrigues
• Journal: Pattern Recognition

• Paper 3: Generalized Multi-Output Gaussian Process Censored Regression (https://doi.org/10.1016/j.patcog.2022.108751)
• Authors: Daniele Gammelli, Kasper Pryds Rolsted, Dario Pacino, Filipe Rodrigues
• Journal: Pattern Recognition

• Paper 4: Predictive and prescriptive performance of bike-sharing demand forecasts for inventory management (https://doi.org/10.1016/j.trc.2022.103571)
• Authors: Daniele Gammelli, Yihua Wang, Dennis Prak, Filipe Rodrigues, Stefan Minner, Francisco Camara Pereira
• Journal: Transportation Research Part C: Emerging Technologies

• Paper 5: Graph Neural Network Reinforcement Learning for Autonomous Mobility-on-Demand Systems (https://doi.org/10.48550/arXiv.2104.11434)
• Authors: Daniele Gammelli, Kaidi Yang, James Harrison, Filipe Rodrigues, Francisco C. Pereira, Marco Pavone
• Journal: 2021 60^th IEEE Conference on Decision and Control (CDC)

• Paper 6: Graph Meta-Reinforcement Learning for Transferable Autonomous Mobility-on-Demand (https://dl.acm.org/doi/10.1145/3534678.3539180)
• Authors: Daniele Gammelli, Kaidi Yang, James Harrison, Filipe Rodrigues, Francisco Pereira, Marco Pavone
• Journal: Proceedings of the 28^th ACM SIGKDD Conference on Knowledge Discovery and Data Mining