Software Open Access

Minimizing the evacuation time of a crowd from a complex building using rescue guides – code

Anton von Schantz

Research code and data used in von Schantz & Ehtamo. Minimizing the evacuation time of a crowd from a complex building using rescue guides. arXiv:2007.00509 [physics.soc-ph]. 2020. (submitted manuscript).

The paper presents a procedure for solving the minimum time evacuation from a complex building using rescue guides, and this repository is its implementation. The crowd is modeled with the physics-inspired agent-based social force model. The solution procedure is a combined numerical simulation and genetic algorithm (GA). The GA iteratively searches for the optimal evacuation plan, while numerical simulations are used to evaluate candidate evacuation plans.

The numerical evacuation simulations are implemented in Python and the GA as Bash scripts that were run on a high performance computing cluster. It should be noted that the procedure is currently computationally very demanding. The repository includes codes for the GA, evacuation simulation and its graphical user interface. For further implementation information, see the associated Github repository

The numerical evacuation simulation codes are based on Jaan Tollander de Balsch's codes and, which he created working as a summer assistant in our research group in Aalto University School of Science, Department of Mathematics and Systems Analysis years 2016 and 2017.

The folders in the repository:
* crowddynamics-simulation contains files for running the GUI
* crowddynamics-qtgui contains the files that build the GUI
* crowddynamics contains all files for simulating the movement of a crowd
* data includes simulation data from running the genetic algorithm
* genetic algorithm includes files to run the genetic algorithm
* simulation files includes files specific for the simulating the conference building and hexagon-shaped area

Note! Some necessary files where missing in the earlier versions. Version 1.2 has all the files needed.

This study was first funded by a grant from the Foundation for Aalto University Science and Technology, and later with a grant from the Finnish Science Foundation for Technology and Economics. The calculations in this study were performed using computer resources within the Aalto University School of Science "Science-IT" project.
Files (7.6 MB)
Name Size
7.6 MB Download
All versions This version
Views 192113
Downloads 2214
Data volume 175.3 MB106.2 MB
Unique views 12586
Unique downloads 2114


Cite as