ESCAPE The project
The evacuation of populations may be the only solution to respond to sudden hazards and reduce risks, which are generally increasing due to rising urbanization along coastlines and in deltas, the presence of high-risk factories in densely populated areas, and climate change. While planning for this strategy is essential, evaluating it outside of a crisis context is challenging.
Computer simulation thus takes over, relying on increasingly realistic models. The goal is to explore the collective effects of thousands of individual decisions made under conditions of stress, imperfect information, and potentially deteriorated environments.
To meet this objective, the models incorporate a wide variety of individual profiles and behaviors, as well as different modes of transportation, including walking, personal vehicles, and public transport. The simulations allow for the calculation of evacuation times, the identification of vulnerable areas within transportation networks, and the highlighting of blind spots in the organization of evacuation plans.
The opportunities for this type of model within crisis management exercise cells are therefore numerous.
This is the major issue that the ESCAPE project aims to address, with the objective of contributing to the design of decision support systems for large-scale evacuations.
The core of our proposal is based on the coupling of geographic information systems (GIS), multi-scale agent-based modeling, and simulation exploration methods. These tools allow us to explore scientific questions that are difficult to address through in-situ observation and, subsequently, to test or explore different evacuation strategies in collaboration with crisis management stakeholders.
Using geographic and demographic information along with a mobility behavior simulator, we establish diagnostics related to evacuation times and identify the local constraints of these evacuations. This work involves exploring the emergence of collective behaviors at different scales that may potentially disrupt the evacuation process, and analyzing possible solutions to mitigate their impact on the evacuation dynamics. The development of demonstrators helps to identify the key issues present in each case study.
The final ESCAPE product will consist of two parts
The first part will be a set of extensions to the GAMA modeling and simulation platform, which will be generic and reusable.
The second part will include all the demonstrators, built using the previously developed extensions.
These two parts will provide the means to define:r :
1. Spatial data (topological structures and land-use patterns), social data (generation and localization of populations at fine scales), and knowledge bases on flow dynamics (counting points, hourly flows
2. Data concerning the hazard, in the form of files from a GIS (hazard maps) or a model simulating the event;
3. Databases on the knowledge and skills of agents according to their roles (citizen, emergency service, crisis manager). This knowledge base will integrate the evacuation plan of the study area, where applicable;
4. A cognitive architecture that will allow for the modeling of complex behaviors related to mobility during a crisis and evacuation management ;
5. The framework for exploring and visualizing evacuation simulations, taking into account different perspectives, at various scales, and according to multiple output indicators
The ESCAPE project is supported by the National Research Agency (ANR-16-CE39-0011-01) and brings together four public laboratories in collaboration with a private risk management entity.
L’Unité Mixte de Recherche « Identité et Différenciation de l'Espace, de l'Environnement et des Sociétés » (UMR 6266 IDEES) – Normandie Université & CNRS) The IDEES UMR brings together around 120 researchers specializing in areas such as spatial modeling and analysis, transportation and port environments, health and risks, Information and Communication Technologies (ICT), and socio-territorial restructuring.
« Laboratoire d'Informatique, de Traitement de l'Information et des Systèmes » (EA 4108) - Université de Rouen Normandie, Université du Havre Normandie et INSA Rouen Normandie.)
LITIS is a research unit in the field of information science and technology in Haute-Normandie. LITIS adopts a multidisciplinary approach, combining practitioners and theorists at the intersection of computer science, pattern recognition, signal and image processing, medicine, and mathematics.
L’Unité Mixte Internationale « Unité de modélisation mathématique et informatique de systèmes complexes » (UMI 209 UMMISCO) – IRD)
UMI Ummisco's research activities focus on mathematical and computer modeling of natural, biological, or social complex systems, and the development of new mathematical and computer modeling methods. These studies have applications in emerging infectious or non-infectious diseases, climate change, natural hazards, and ecosystems and natural resources.
« Institut des Systèmes Complexes de Paris Ile de France » L'ISC-PIF is a CNRS research and service unit dedicated to the inter-institutional and interdisciplinary development of complex systems research. It serves as a research laboratory, project incubator, shared resource center, conference center, and academic co-working space. This scientific hub provides researchers with a dynamic research environment and innovative tools based on big data and high-performance computing.
La société d’ingénierie Bas Rhône Languedoc Ingénierie (BRLi) A private company specializing in crisis prevention and management, BRLi operates in more than 80 countries, with a significant focus on Sub-Saharan Africa, the Mediterranean, and the Middle East. Two-thirds of BRLi’s activity is driven by water-related services, including integrated water resource management and water and risk management.
