Section: Application Domains

Transport Systems

Participants : Stefan Haar, Serge Haddad, Yann Duplouy, Simon Theissing.

We participate in the IRT System X’s system of systems program TMM, in two projects:

  • project MIC (terminated in November 2016) on multi-modal transport systems with academic partners UPMC, IFSTTAR and CEA, and several industrial partners including Alstom (project leader), COSMO and Renault. Transportation operators in an urban area need to plan, supervise and steer different means of transportation with respect to several criteria:

    • Maximize capacity;

    • guarantee punctuality and robustness of service;

    • minimize energy consumption.

    The systems must achieve these objectives not only under ideal conditions, but also be robust to perturbations (such as a major cultural or sport event creating additional traffic), modifications of routes (roadwork, accidents, demonstrations, ... ) and tolerant to technical failures. Therefore, systems must be enabled to raise appropriate alarms upon detection of anomalies, diagnose the type of anomaly and select the appropriate response. While the above challenges belong already to the tasks of individual operators in the unimodal setting, the rise of and increasing demand for multi- modal transports forces to achieve these planning, optimization and control goals not in isolation, but in a cooperative manner, across several operators. The research task here is first to analyze the transportation system regarding the available means, capacities and structures, and so as to identify the impacting factors and interdependencies of the system variables. Based on this analysis, the task is to derive and implement robust planning, with tolerance to technical faults; diagnosis and control strategies that are optimal under several, possibly different, criteria (average case vs worst case performance, energy efficiency, etc.) and allow to adapt to changes e.g. from nominal mode to reduced mode, sensor failures, etc.

  • the project SVA ( Simulation pour la Sécurité du Véhicule Autonome ), where the PhD Thesis of Yann Duplouy targets the application of formal methods to the development of embedded systems for autonomous vehicles.