Ifsttar PhD subject

 

French version

Detailed form :

Title : Data-driven and model-driven hybrid modeling for the evaluation of shared mobility systems

Main host Laboratory - Referent Advisor COSYS - GRETTIA  -  ZARGAYOUNA Mahdi      tél. : +33 181668698 
Director of the main host Laboratory OUKHELLOU Latifa  -  
PhD Speciality Transport, informatique, mathématiques appliquées
Axis of the performance contract 1 - COP2017 - Efficient transport and safe travel
Main location Marne-la-Vallée
Doctoral affiliation UNIVERSITE GUSTAVE EIFFEL
PhD school MATHEMATIQUES ET SCIENCES ET TECHNOLOGIES DE L'INFORMATION ET DE LA COMMUNICATION (MSTIC)
Planned PhD supervisor ZARGAYOUNA Mahdi  -  Université Gustave Eiffel  -  COSYS - GRETTIA
Planned PhD co-supervisor OUKHELLOU Latifa  -  Université Gustave Eiffel  -  COSYS - GRETTIA
Planned financing Contrat doctoral  - Ifsttar

Abstract

Background
The increasing pressure of demand on urban transportation systems requires innovative solutions that can increase its efficiency. In recent years, intelligent transportation systems have reshaped the traditional transportation offer with the rapid introduction of new mobility services. Mobility as a service (MaaS) and shared mobility concepts can contribute to the creation of healthier, cleaner and more accessible cities. Shared mobility includes all transport services that can be shared by users. They include public transportation, cabs, carpooling, car-sharing, bicycle and scooter sharing, etc.
There is currently a clear trend towards the implementation of automated cooperative mobility services; although the expected timeframes, technological options and use cases remain uncertain, decision-makers must already prepare their responses to these developments. The integration of shared use mobility services, particularly in large urban areas, follows a parallel trend to that of shared vehicle ownership. The potential of mobility solutions such as carpooling and transportation-on-demand to meet urban transportation demand is attracting increasing attention.
Moreover, in multimodal trips, the first and last mile are crucial. Conventional public transport is, in most cases, unable to provide first- and last-mile transportation, particularly at times of low demand and in low-density locations [1]. In particular, concepts of shared mobility and vehicle automation have the potential to radically improve the mobility service, allowing a paradigm shift in urban mobility [2]. From the system perspective, the co-existence of shared mobility systems and traditional public transport system leads to diversified and complex systems that enable more instruments to alleviate traffic congestion and improve users’ quality of life, provided that the comprehensive system of systems can be designed, modeled and operated effectively [3]. In other words, the model should be able to represent the actual performance of the mobility service and its interaction with the other modes of the system.
Objective
The main objective of this thesis is to study and develop models to represent, manage and optimize new mobility modes and services, in an integrated way with the public transport system. These models will allow in a second step to capture and compute the current state of the network (user equilibrium) and the targeted state of the system (system optimum) in an integrated mobility system.
Methodology
The contributions of this thesis will be based on data mining on the one hand and on multimodal traffic simulation (design, calibration and validation) on the other hand. Indeed, the mobility system being very complex by nature, the modeling should take advantage of the strengths of both approaches. On the one hand, data-oriented modeling through data mining will allow to build robust systems with high descriptive and predictive power and good speed of execution. On the other hand, model-oriented approaches allow to simulate situations that have never been observed in the system data. In addition, the optimization system will be partly based on clustering approaches, enabling good quality solutions to be found in very short execution times. These approaches will indeed allow describing the system by a reduced number of parameters while keeping the amount of information necessary for its fine description, thus inducing low calculation times for the search of optimal solutions.
As far as calculating the user equilibrium and the system optimum is concerned, it differs according to the application scenario. Indeed, service providers can work in competition or in cooperation. In the case of cooperative interaction, they share fleet information to facilitate passenger transportation, and the benefits of carpooling will be more easily achievable. In the competitive scenario, each provider tries to achieve a higher profit. The matching algorithm communicates with the suppliers directly or indirectly. The user equilibrium and system optimum would be different for each category. We will work in the thesis to implement and compare these different strategies while positioning them in the literature on this subject.
Existing tools and models
GRETTIA has a long experience with the modelling of shared mobility [4][5][6][7]. This thesis will notably be based on the optimization system designed and implemented in the framework of Negin Alisoltani's thesis, recently defended [8]. The laboratory also has a recognized expertise in data-oriented approaches applied to urban mobility [9,10]. Finally, the laboratory is actively working on allocation models and traffic equilibrium models [11, 12]. This thesis will seek to take advantage of each of the laboratory's advances to propose a new methodological framework at the intersection of operational research and machine learning.
Candidate profile
The candidate must:
1) Have an Master 2 or equivalent in transportation engineering, civil engineering, computer science, urban planning, operations research or other field strongly related to transportation.
2) Have excellent analytical and communication skills in written and spoken English.
3) Be able to work independently and take responsibility for the progress and quality of the project.
4) Have experience in traffic data collection, statistical data analysis and exploration, and geospatial data analysis.
5) Have very good programming skills.
References
[1] Boarnet, M. G., Giuliano, G., Hou, Y., & Shin, E. J. (2017). First/last mile transit access as an equity planning issue. Transportation Research Part A: Policy and Practice, 103, 296-310.
[2] Pinto HK, Hyland MF, Mahmassani HS, Verbas IÖ. Joint design of multimodal transit networks and shared autonomous mobility fleets. Transportation Research Part C: Emerging Technologies. 2019 Jun 22.
[3] Pi X, Ma W, Qian ZS. A general formulation for multi-modal dynamic traffic assignment considering multi-class vehicles, public transit and parking. Transportation Research Part C: Emerging Technologies. 2019 Jul 1;104:369-89.
[4] Alisoltani N, Zargayouna M, Leclercq L. A Sequential Clustering Method for the Taxi-Dispatching Problem Considering Traffic Dynamics. IEEE Intelligent Transportation Systems Magazine. 2020 Sep 9;12 (4):169-81.
[5] M Zargayouna, B Zeddini, Dispatching Requests for Agent-Based Online Vehicle Routing Problems with Time Windows, CIT. Journal of Computing and Information Technology 28 (1), 59—72
[6] M Zargayouna, B Zeddini, Fleet organization models for online vehicle routing problems, Transactions on Computational Collective Intelligence VII, 82-102
[7] F Grootenboers, M De Weerdt, M Zargayouna, Impact of competition on quality of service in demand responsive transit, Lecture Notes in Computer Science 6251, 113-124
[8] Negin Alisoltani, Frameworks d'optimisation basée sur la simulation pour le covoiturage dynamique, thèse de doctorat, Université de Lyon, 2020
[9] Etienne Côme, Latifa Oukhellou, Model-based count series clustering for bike sharing system usage mining: a case study with the Vélib’system of Paris, ACM Transactions on Intelligent Systems and Technology (TIST) 5 (3), 1-21
[10] K Mohamed, E Côme, L Oukhellou, M Verleysen, Clustering smart card data for urban mobility analysis, IEEE Transactions on intelligent transportation systems 18 (3), 712-728
[11] Ameli M, Lebacque J. P, Leclercq L. Simulation-based dynamic traffic assignment: meta-heuristic solution methods with parallel computing. Computer-Aided Civil and Infrastructure Engineering, 2020, 35(10), 1047-1062.
[12] Ameli M, Lebacque J. P, Leclercq L .Cross-comparison of convergence algorithms to solve trip-based dynamic traffic assignment problems. Computer-Aided Civil and Infrastructure Engineering, 2020, 35:219–240.

Keywords : New mobility services, simulation, data analysis, traffic assignment, multimodal traffic
List of topics
Applications closed