Team : Flaming Hot Flamingos
Member : Jessica C.Y. Lee (AUS) Mohamed Nabil bin Mohamed Hairol (BRN) Chanyoung Lee (ROK)
Mentor : Prof. Neil Satoha
Research Title : Fighting Natural Disasters: Improving the Effectiveness of Flood Early Warning Systems with AI and Reducing the Areas of Impact

Research summary :
Members of the APEC region often experience natural catastrophes due to where their country is positioned and the ever changing climate on earth as global warming occurs. With the ongoing advances in artificial intelligence, machine learning-based flood forecasting systems have emerged, offering the potential for higher accuracy and faster response. In this study, data was collected on AI-enhanced Early Warning Systems and Traditional Early Warning Systems for floods. The datasets were then compared to gauge the effectiveness of AI-enhanced Early Warning Systems in countries with varying levels of AI adoption.

For this case study the independent variable was the type of forecasting infrastructure and we measured the accuracy and lead times of warnings given by forecasting systems. The data on AI-based Flood Forecasting Systems was collected by accessing public platforms that provide real-time AI flood predictions. Whilst the data on traditional forecasting systems was obtained by looking at systems using sensor-based hydrological models or manual disaster agency alerts. Then, a performance comparison between the two types of systems were made. From the data collected so far, AI significantly improves on the current Traditional Early Warning Systems on effectiveness, speed and validity of alerts. This research contributes to climate resilience by highlighting the role of AI in improving a system’s ability to predict and respond to flood disasters.