Floodify
/Floodify

Floodify

Tackling urban flash floods through real-time data integration

Smart cities logo

A case for climate adaptation: urban flash floods

Due to a changing climate, extreme weather events such as flash floods are becoming more common. Their destructive impact is most severe in urban environments and calls for urgent climate adaptation. This requires better prediction and impact estimation, and ultimately better urban planning.

Flood forecasting has already improved, using technologies such as remote sensing and social media monitoring. However, these methods still lack the real-time responsiveness and accuracy needed for precise predictions. They also fall short in terms of integrating real-time, heterogeneous data sources into a cohesive system. Data is still often siloed, preventing timely sharing between key stakeholders. Also, many current models are computationally heavy, making them impractical for real-time use, especially in fast-evolving flash flood scenarios.

Innovative flash flood prediction

Floodify envisages an innovative approach to flash flood prediction by developing a real-time, data-driven system that leverages heterogeneous data from multiple sources.

The project's core innovation is the creation of a Flash Flood Data Space, a decentralized infrastructure where various stakeholders—ranging from urban planners and emergency services to data providers—can share and access data in near real-time. This data space will integrate inputs such as sensor data, drone imagery, social media reports, satellite observations, and more, creating a rich, multi-dimensional view of urban flash flood risks.

Floodify’s solution will go beyond existing approaches by using advanced data fusion and image processing techniques. These technologies will combine disparate data types and sources to generate accurate and timely flood models. Real-time data updates will allow for immediate identification of flood-prone areas, facilitating faster emergency responses and more proactive urban planning.

Research lines

Floodify’s solution is built around three research lines, each focused on a different aspect of the flash flood prediction challenge:

  1. Flash Flood Data Space:

This research line will develop the technical and functional architecture of the Flash Flood Data Space. By studying various data sources—such as drone images, sensor networks, and citizen reports—the project will define how to best share this data in a secure, interoperable, and near real-time manner. The aim is to create a decentralized platform where data from multiple actors can be shared and fused to improve flash flood forecasting.

  1. Image Processing and Data Fusion:

Floodify will develop advanced image processing techniques to extract flood information from aerial and terrestrial images. Additionally, it will design data fusion algorithms that combine data from diverse sources with varying quality and geographical distribution, producing highly accurate flood severity predictions.

  1. Data Assimilation for Flood Modeling:

The third research line focuses on integrating real-time data from the Flash Flood Data Space and image processing outputs into existing flood models. By assimilating this new data, the project will refine flood prediction models, significantly improving their granularity and reducing uncertainties. This will enable faster, more precise flood forecasting and impact assessments, particularly for flash floods.

“Floodify’s real-time, integrated system promises to reduce the uncertainty in flash flood forecasting by 50%, giving decision-makers the tools they need to mitigate risks and save lives.”

Floodify

Floodify develops a scalable, real-time solution to urban flood management.

Floodify is an imec.icon research project funded by imec and Agentschap Innoveren & Ondernemen (VLAIO).

The project started on 01.09.2024 and is set to run until 28.02.2027.

Project information

Industry

  • Aquafin
  • Brandweerzone Centrum
  • CityMesh
  • G.I.M – Geographic Information Management (GIM)
  • HydroScan
  • MyCSN

Research

  • imec – EDiT – UGent
  • imec – IPI – UGent

Contact

  • Project lead: Jan Geukens
  • Research lead: Tanguy Coenen
  • Proposal manager: Maxim Chantillon
  • Innovation manager: Deben Lamon