RAPID EARTHQUAKE LOSS ASSESSMENT FRAMEWORK: RELAR PROJECT AND ITS APPLICATION ON THE DATA SET FROM THE KRALJEVO (SERBIA) EARTHQUAKE IN 2010

Rapid earthquake damage and loss assessment is a critical process for evaluating the immediate impacts of seismic events on infrastructure and populations. It involves the quick analysis of structural damage and economic losses. The goal is to provide timely information to emergency responders, government agencies, and decision-makers, enabling efficient resource allocation and disaster response. Advanced technologies such as machine learning, remote sensing, and real-time data analytics have improved the accuracy and speed of these assessments, helping to mitigate the effects of earthquakes and support recovery efforts. This paper presents the RELAR project, funded by the Science Fund of the Republic of Serbia, which aims to improve earthquake loss assessment and recovery processes. By integrating Machine Learning and Image Recognition, the project accelerates response times and enhances the accuracy of damage estimation and repair cost assessments. Traditional methods often suffer from delays and inaccuracies due to data limitations and lack of flexibility. RELAR offers innovative solutions for providing reliable, timely information, even in the absence of ground motion data. The project’s objectives include developing practical ML algorithms, validating assessment models, and establishing proactive risk mitigation strategies. Within the paper, an application to the case of Kraljevo 2010 earthquake is presented, showing the methodology of the approach.