[DFG/ANR project MADRAS]
Multi-agent modelling of dense crowd dynamics:
Predict & Understand
PhD opening on physics-based models for pedestrian motion at University de Lyon 1
Required expertises: Statistical Physics, Computer programming
We are looking for a bright and motivated candidate for a doctoral position (PhD) on the physics-based modelling of pedestrian motion, at Institut Lumière Matière, in Lyon. The PhD may be preceded by a Masters' internship.
In order to design new public venues and dimension them prior to their construction or to have autonomous cars adapt to pedestrians around them, it is critical to be able to predict and anticipate the motion of pedestrians in the midst of crowds. Unfortunately, currently used models suffer from severe deficiencies, especially at high densities [1-3].
Over the last years, research works have demonstrated the relevance of a key variable, namely, the anticipated time to collision, in controlling interactions between pedestrians . Thus, a novel branch of models have started to emerge, in which pedestrians are no longer treated a driven grains, but as agents capable of choosing their trajectory so as to minimise a given 'energy'. Not only has this change of paradigm raised physical questions, but it has only given rise to technical challenges for the implementation of the models. For instance, only cylindrical pedestrians can be modelled so far.
The PhD, which may be preceded by a research internship, will aim to unlock the situation by developing and testing computationally efficient models based on these ideas. It will thus target - the resolution of the conceptual questions raised by the new framework - the definition of efficient implementation strategies - the calibration and testing of the developed programme, by comparison with empirical data - the exploitation of the model to simulate actual mass gatherings.
The PhD is funded in the frame of French-German project MADRAS (subsidised by ANR and DFG), which gathers researchers from the Research Center Jülich (Mohcine CHRAIBI), Wuppertal University (Antoine TORDEUX), the Université de Toulouse (Benoit GAUDOU, Nicolas VERSTAEVEL) as well as the University of Lyon.
The ideal candidate will hold a Masters degree (or an Engineers' degree) in Statistical Physics, Applied Mathematics, Computer Science, or Fluid Mechanics, and (s)he will be eager to be involved in an interdisciplinary project.
He/she has strong skills in numerical programming (C++, Python) and an inclination for exploratory research. He/she will hold a degree in Statistical Physics or Applied Mathematics. He/she will be based at iLM (Lyon 1 / Villeurbanne).
This position is fully funded by the Agence Nationale de la Recherche within the French-German project MADRAS, whose exciting goal is to develop, validate, and exploit novel models to describe dense crowd dynamics.
The application process is now open. Candidates are expected to send their CV by email along with a cover letter specifying their scientific interests and any relevant details about their curriculum. Informal enquiries are also welcome.
 MOUSSAÏD, Mehdi, HELBING, Dirk, et THERAULAZ, Guy. How simple rules determine pedestrian behavior and crowd disasters. Proceedings of the National Academy of Sciences, 2011, vol. 108, no 17, p. 6884-6888.
 KARAMOUZAS, Ioannis, SKINNER, Brian, et GUY, Stephen J. Universal power law governing pedestrian interactions. Physical review letters, 2014, vol. 113, no 23, p. 238701.
 NICOLAS, Alexandre, KUPERMAN, Marcelo, IBAÑEZ, Santiago, et al. Mechanical response of dense pedestrian crowds to the crossing of intruders. Scientific reports, 2019, vol. 9, no 1, p. 105.