Mohcine Chraibi

Head of the division Pedestrian Dynamics - Modeling at Civil Safety Research

About Me

Hi, I’m Mohcine, a senior researcher at Forschungszentrum Jülich focusing on pedestrian dynamics and crowd behavior. My research combines computational modeling with empirical validation to better understand how people move and interact in crowds. Want to see what I do in action? Check out this beautiful comic
Bustling through the physics of crowds.

I’m passionate about open source software for simulating, analyzing and visualizing pedestrian dynamics, which I believe is fundamental to good scientific practice.

To help grow our research community, I maintain a website that collects useful information and news about pedestrian dynamics at pedestriandynamics.org and a GitHub-organisation at github.com/PedestrianDynamics.


Research interests: Mathematical ModellingSimulationData AnalysisResearch Software Development

Projects

Multi-agent modelling of dense crowd dynamics: Predict & Understand

Project Website

The French-German project MADRAS

The goal of this joint-project is to develop models to predict and understand dense crowd movements of 2-8 people per square meter. The research combines neural networks with physics-based modeling to improve upon current models that struggle with high-density scenarios. The project integrates these approaches into an online platform for real-time visualization and comparison with actual crowd data. Using the Festival of Lights as a case study.

JuPedSim: Flexible and Expressive Pedestrian Simulation

Project Website

Open-source Python package with a C++ core to simulate pedestrian dynamics.

JuPedSim bridges core functionality with user interaction through a lightweight and expressive Python API. This interface enables users to define simulation scenarios, prototype complex behaviors, and leverage core components with ease. The latest version transforms JuPedSim into a Python-based library, offering greater flexibility in simulation configuration and integration.

PedPy: Intuitive and Versatile Pedestrian Analysis

Project Website

PedPy: Pedestrian Trajectory Analyzer

PedPy is an open-source, MIT-licensed Python library designed for pedestrian movement analysis. It offers a high-level interface to extract fundamental measurements from trajectory data, enabling direct computation of velocity, flow, and density. These can be combined to generate fundamental diagrams, while additional functionalities allow for the computation of density and velocity profiles.

Walk the Line: The role of gender and culture

Project Website

Streamlit application to visualize experiments performed in five countries.

For the first time, we explore external validity and cross-cultural generalizability by conducting the same experiment in five different countries. Each study examines the impact of varying gender compositions on single-file pedestrian movement. Our findings reveal consistent effects of gender composition across all locations, with no significant deviations in the fundamental diagrams, particularly in the bounded regime.

Experience

Research Centre Jülich

Civil Safety Research (IAS-7)

November 2018 - Present

Division Head

How can we understand the properties of a crowd and predict its dynamics in closed and open spaces?

I tackle this question head-on, developing operational models and investigating route choice models in the field of software engineering. My research is centred around developing JuPedSim, an open-source software for simulating pedestrian movement.

University of Wuppertal

School of Architecture and Civil Engineering

2010 - 2012

Research associate

Education and research

As an employee and lecturer at Bergische Universität Wuppertal, I contributed to teaching through my lecture informatics, where civil engineering students were introduced to numerical engineering methods and programming.

Research Centre Jülich

Jülich Supercomputing Centre

November 2008 - 2018

Phd and postdoc

Boundary effects and non=equilibrium states in pedestrian dynamics - experiments and modelling

I worked as a postdoc on several projects, including kapaKrit, where we studied how to maximize evacuation capacity in the short term through structural, technical, and organizational measures, using Dortmund Central Station as a case study. To achieve this, we modeled and simulated the complex train station system to identify key parameters for capacity optimization.

Education

University of Wuppertal

Habilitation Computer-aided Traffic Simulation

2024

Elements for modeling pedestrian movement from theory to application and back

In my habilitation defense I delivered two lectures. In my first presentation, I introduced car-following models, their historical evolution, and stability characteristics, using NetLogo simulations to facilitate student understanding. My second lecture showcased a numerical reconstruction of the 1919 Amritsar massacre using JuPedSim, demonstrating how modeling techniques can provide new perspectives on historical events.

This achievement grants me the authority to independently teach and supervise PhD students.

University of Cologne

PhD thesis

2008 - 2012

Validated force-based modeling of pedestrian dynamics

Investigation of force-based modeling of pedestrian dynamics. Focusing on the quantitative validation of mathematical models, this work addresses fundamental questions: Is it possible to describe pedestrian dynamics solely using equations of motion derived from Newtonian dynamics?

Technical University Hamburg-Harburg

Diploma thesis

2002 - 2008

Systems of Ordinary Differential Equations for Describing Pedestrian Dynamics

The aim of this thesis is to examine the realism of models concerning pedestrian dynamics and to analyze challenges in their numerical computation. The focus of these investigations is the centrifugal force model. Based on the findings, I propose approaches to improve the modeling accuracy, particularly with regard to numerical simulation.

Teachings at the University of Wuppertal

Pedestrian Dynamics

English

Winter semeser

How can we understand the properties of a crowd and predict its dynamics in closed and open spaces?

This lecture addresses this question that weaves together three streams: development of operative models, investigation of route choice models, and software engineering. In this lecture, we aim at exploring several interrelated topics that fall under “mathematical modeling of pedestrian dynamics.

Traffic Flow Modeling

German

Winter semeser

Fundamentals of Modeling Car-Following Models with NetLogo

This lecture covers key concepts in traffic flow modeling, including modeling scales, model types, and traffic simulation tools. It explores cellular automata and car-following models, tracing their historical development and advancements. Additionally, force-based models for pedestrian dynamics are introduced. Hands-on exercises and demonstrations using NetLogo provide practical insights into simulation techniques.

Evacuation simulation

German

Summer semeser

Evacuation simulations covering pedestrian movement models, software application

This lecture provides an introduction to evacuation simulations. It is designed for master’s students in safety engineering and civil engineering. The lecture consists of lectures and practical computer-based exercises, offering insights into models for pedestrian and crowd movement in buildings and other structures. Additionally, students learn to use the software (JuPedSim and Pathfinder) and critically interpret simulation results. At the end of the course, the acquired methods are applied in a project, such as simulating the evacuation of a building or analyzing pedestrian flows at an event.