About

What is ARTIFACT?

ARTIFACT (ARTificial Intelligence for Flood Resilient Infrastructure) is an ambitious Horizon Europe, Twinning Green Deal project aimed at addressing the increasing threats posed by urban flooding due to climate change and urbanization. By combining cutting-edge artificial intelligence (AI) with nature-based solutions (NBS), ARTIFACT seeks to revolutionize flood modeling, forecasting, and sustainable infrastructure planning.

The Institute for Artificial Intelligence Research and Development of Serbia (IVI) is leading this initiative, emphasising the importance of integrating nature-based solutions and AI in urban planning to address future flood risks. IVI aims to use large data sets and advanced machine learning techniques to analyse climate change impacts. Additionally, the project will establish a Blue-Green AI Hub at IVI to connect research with the regional information and communication technology (ICT) industry. This will all be done with the help of EU project partners Delft University of Technology (TU Delft), Delft Institute for Water Education (IHE Delft), and Hamburg University of Technology (TUHH).

Key Objectives

Enhancing Research Excellence: Developing AI-based flood forecasting and integrating NBS technologies.
Innovation and Capacity Building: Strengthening the capabilities of the Institute for Artificial Intelligence Research and Development of Serbia (IVI) in collaboration with leading EU institutions.
Creating a Knowledge Hub: Establishing the Blue-Green AI Hub to foster innovation and connect researchers with industry and policymakers.
Empowering Researchers: Training a new generation of AI researchers specialized in urban flood resilience.

The need for ARTIFACT

Climate change and urbanization have exacerbated the risks and impacts of urban flooding, leading to billions of euros in damages annually across Europe.

Traditional flood defense systems are expensive and space-consuming, highlighting the need for innovative, distributed solutions like NBS.

Current flood models lack real-time adaptability and efficiency, limiting their utility in early warning systems..