From Data to Decisions: An AI- and UAV-Driven Pipeline for Bushfire Management

13 Ottobre 2026
Share on facebook Share on twitter
56
VIEW

Oct 13 h 2:30 PM
DII Meeting Room - 6th floor
School of Engineering, Largo L Lazzarino 1 Pisa

Jahan Hassan, School of Engineering and Technology at Central Queensland University (CQU), Australia. 

 

Presentation abstract: Wildfires are a growing threat across many parts of the world, and Australia and Italy, despite their different landscapes, share common challenges: lengthening fire seasons, increasingly unpredictable fire behaviour, and the need for tools that support decision-making across the full fire management lifecycle. This talk presents a line of research, developed in the Australian context but broadly transferable, that combines remote sensing, machine learning, and UAV simulation across five interconnected stages of that lifecycle. We begin at the regional scale, using twelve years of satellite imagery and an XGBoost model to map fire severity and predict future risk across Australia. We then examine how data quality and quantity shape detection algorithm performance, before turning to the UAV itself: transfer-learning-enhanced YOLO models for safe obstacle-aware navigation in forest environments. Once a fire is detected, we present real-time methods for estimating fire area from UAV imagery, and a simulation-to-learning framework using an encoder-decoder LSTM to forecast fire spread. The talk closes by reflecting on shared challenges and directions for future research across fire-prone regions.

Presenter Bio: Associate Professor Jahan Hassan is an academic in the School of Engineering and Technology at Central Queensland University (CQU), Australia. She holds a PhD in Computer Science from the University of New South Wales (UNSW) and a Bachelor of Computing from Monash University, following prior academic and research roles at UNSW, the University of Sydney, and the Smart Internet Technology Cooperative Research Centre. Her research focuses on UAV-assisted wireless networks, sensing technologies, and the Internet of Things (IoT), with applications in smart agriculture, environmental monitoring, and natural hazard prediction. She currently leads a funded project on AI- and drone-assisted targeted weed management and supervises research projects across UAV applications, machine learning, and bushfire management. A recipient of CQU's Dean's Award for Research Excellence, Associate Professor Hassan serves as an Editor for Ad Hoc Networks (Elsevier), a Senior Member of IEEE, and co-chair of the IEEE WoWMoM DroneSense-AI workshop series. Her educational leadership has also been recognized nationally with the 2024 Australian Award for University Teaching (AAUT) and multiple Vice-Chancellor's awards.