Mieux protéger les populations vulnérables face aux inondations

Every year, severe floods affect millions of vulnerable people in developing countries where the lack of data and infrastructure make rescue missions especially complex. Developing an innovative tool would help decision-makers to determine the best locations for permanent shelters and thus limit the risk to evacuated populations.

Establish Strategic Locations

Developed in cooperation with the World Bank and the Haitian government, this optimization tool incorporates flood maps, demographics, road networks, and vulnerability indicators to determine automatically the most strategic locations to maximize the safety of at-risk populations while respecting budget and logistics constraints.

Validated with data from the Nippes Department in Haiti to test various scenarios of shelter locations based on community service level and risk exposure, then expanded to include the North West Department; the model also takes into account an often-overlooked factor: water level and its impact on travelling speed on foot, an undeniable reality of rural areas with no vehicle access.

Maximize Safety

Affected communities will benefit directly: rapid access to safe shelters will reduce their exposition to difficult and dangerous conditions; enhanced protection of their dignity; and increased resilience to cope with future disasters.

This project also represents a significant step forward in operations research applied to humanitarian issues by:

• Producing optimal results in mere seconds and promptly adapting to evolving situations.
• Providing better solutions than traditional approaches based on spatial assessments and field trips.
• Determining the location of shelters targeting high-risk areas more effectively and suggesting safer evacuation routes.
• Providing a flexible tool applicable to contexts where historical data are rare, such as developing countries and remote rural areas in industrialized countries.

Innovate to Improve Well-Being

Équipe de recherche

Maedeh Sharbaf, Ph.D. student; Valérie Bélanger and Marie-Ève Rancourt, Associate Professors, Department of Logistics and Operations Management, HEC Montréal; Marilène Cherkesly, Professor, Department of Analytics, Operations, and Information Technologies, UQAM; Giovanni Michele Toglia, Systems Engineer, World Bank