Actualité - Annonce de Thèse/HDR

Date : July 10, 2024, 2:30 p.m. - Type : Thesis - Jean JODEAU - Amphithéatre Campus Charpak

Optimization and Simulation of a rail-road transportation system in sparsely populated area

In France, sparsely populated areas suffer from a lack of mobility. To address this, SNCF has developed a small 14-seater vehicle. It can travel on the road network and on a dedicated rail network. The use of these vehicles will revitalise a currently unused rail network, while providing a new mobility solution for residents. The work of this thesis takes place in this context.


The operation of a rail-road vehicle in commercial services is new for  SNCF. The aim of this work is to design the operating algorithms for a fleet of these vehicles. Two operating systems have been studied in our work: the first offers great flexibility of use for users in the form of a door-to-door on-demand transportation system, while the second is easy to implement for the operator and is based on the design of timetabled lines.

Our work has therefore focused on the study of these two transport systems and the development of operational algorithms that can consider the flexibility offered by road transport and the mobility offered by the rail network, while respecting the constraints specific to the rail world.


First, we present a mathematical model to define the Dial-A-Ride problem with Railway Constraints. We then present a heuristic solution method that makes it easy to introduce railway constraints when designing a transport plan for customers who have booked.  This method is then adapted to include customers on the fly and to adjust vehicle routes to best serve these new customers. For each method, we present an evaluation of its performance using indicators from the literature as well as indicators dedicated to this new vehicle.


Secondly, we define and present the problem of designing timetabled services around a railway line that serves as the backbone of the service. Performance is then evaluated using criteria from the literature and adapted to a rail-road service.


In the third and final stage, we integrate the on-demand transport algorithms into a simulator to evaluate precisely and under as realistic conditions as possible the performance of such a system in the target area. We also present a comparison of the performance of the two transport systems in the area chosen by SNCF.



  • PELLEGRINI Paola, Directrice de Recherche, UGE (Rapporteure)
  • VERCRAENE Samuel, Professeur Associé, INSA Lyon (Rapporteur)
  • CATTARUZZA Diego, Professeur, Ecole Centrale de Lille (Examinateur)
  • ELLOUMI Sourour, Professeure, ENSTA Paris (Examinatrice)
  • PRODHON Caroline, Professeure, Université Technologique de Troie (Examinatrice)
  • FEILLET Dominique, Professeur, Mines Saint-Etienne (Directeur de thèse)
  • ABSI Nabil, Professeur, Mines Saint-Etienne (Co-Directeur de thèse)
  • CHEVRIER Rémy, PhD, Chef de projet MOD, SNCF (Co-encadrant).