Decision Support System for Robust Design and Service Quality Indicators of Railway Transport Plans – SCRIPT

The first step of the project has been to make a state of the art of the various indicators and methods proposed in literature to assess stability in the railway field, but also to identify the interactions between different critical rail resources. Combined with this work, various discussions with SNCF planners have permitted to build a common view among all the partners involved in the project. In addition, a new specific user-opinion pool has been conducted for the project. All these elements should permit to select several relevant indicators to quantify the robustness of railway transport plans.
Simultaneously to the work on robustness indicators, we have developed a software prototype which can generate railway transport plans and is based on an integrated optimization of the various resources. A flow simulation model has then been developed. This model allows to reflect the operational behavior of a transport plan under various disturbances and thus to quantify the impact of each possible disturbance. Experiment has been performed on real railway data in order to validate both prototypes.

The main result of the project was to show that designing transport plans in an integrated way can help to significantly improve their robustness. Moreover most of the improvements observed were obtained for a constant cost of transport.
From a scientific point of view, the robustness of railway plans was only studied for each resource separately and our innovative approach could initiate many researches.

Future works should focus on three main points:
- From an algorithmic point of view, the optimization method developed could certainly be improved. In addition, some potential bias have been highlighted and should be corrected.
- Concerning the optimization criteria, it could be interesting to consider one related to the reparability of railway transport plans in order to evolve toward a proactive-reactive approach.
- Finally, in terms of decision support, it would be useful to determine the correlations between the different optimization criteria and the quantification of each disturbance impact in order to set the importance of criteria.

Researches from the project have generated two articles published in the proceedings of international conferences, two talks in national conferences and a PhD thesis.

The SCRIPT project aims at characterizing and answering robustness problems in the design of railway transport plans. The first objective will be to define the relevant indicators for modeling the stability of a plan from the point of view of the users and customers as well as from the point of view of the railway company. The project then aims at proposing innovating methods that determine optimized stable plans compared to a typology of risks, and at validating these methods with a simulator integrating the various risks on real data. This project falls in the thematic axes 3.2 (“gestion des réseaux de transport”, i.e. management of transport networks) and 3.3 (“productivité et services”, i.e. productivity and services) of the call for projects for 2012 of the TDM program. The originalities of the project are multiple. First, the goal is to analyze stability with the double vision customer, mainly concerned with quality of service, and Railway Company, whose primary focus is on costs. The study will consider French customers, whose expectations can be quite different from those of customers in other countries. Secondly, all the critical resources necessary to implement a transport plan will be considered: Train paths, rolling stock and train drivers. A complete network will be considered with its connections with other networks. Lastly, to undertake this research, the project lies on a consortium made up of two complementary laboratories on the problems of robustness, optimization in transportation and simulation, and the main actor of the rail transport in France.