Document Type : Original Research Paper
Authors
1 Department of Remote Sensing and GIS, Faculty of Geography, Tehran University, Tehran, Iran
2 Department of Surveying Engineering, Faculty of Civil, Water and Environmental Engineering, Shahid Beheshti University, Tehran, Iran
Abstract
Background and Objectives: Earthquakes, among the most unpredictable and devastating natural disasters, result in significant human casualties and financial losses worldwide each year. Their sudden occurrence and destructive potential categorize them as critical crises that demand efficient and innovative management strategies. Contemporary crisis management practices emphasize three key phases: preparedness (before the event), response (during the event), and recovery (after the event). Among these, rescue operations, which are part of the response phase, play a vital role in saving lives and mitigating further damage. However, given the urgency and complexity of rescue efforts, enhancing their effectiveness through innovative methods is essential. This study introduces a novel approach that leverages spatial intelligence—specifically Geo-Artificial Intelligence (Geo-AI)—to optimize rescue operations in the aftermath of an earthquake.
Methods: This research proposes a Geo-AI–based framework to enhance rescue performance following an earthquake. The approach involves simulating a hypothetical earthquake scenario in Tehran using the Japan International Cooperation Agency (JICA) floating scenario model. A total of 48 rescuers are organized into six teams within the designated study area. These teams are tasked with conducting search and rescue missions facilitated by an augmented intelligent spatial information system. Unlike traditional or manually assigned rescue operations, the proposed model employs reinforcement learning—a subfield of artificial intelligence—to dynamically allocate resources and optimize operational decisions in real-time. The design incorporates a comprehensive set of variables known to influence post-earthquake rescue effectiveness, including team location, response time, victim survivability, and route accessibility. The aim is to minimize response times and maximize the number of successful rescues using spatially informed decision-making.
Findings: Due to the inherent unpredictability of earthquakes and the logistical constraints of studying rescue operations in real-world post-disaster settings, this research relies on simulation to replicate realistic conditions. The simulation environment provides detailed spatial and descriptive information regarding both the affected area and the status of rescue teams. Furthermore, it enables estimation of structural and human damage resulting from the hypothetical earthquake. Based on these simulated conditions, rescue operations are prioritized according to urgency and resource availability. All 48 rescuers are initially positioned at the nearest crisis management center and are subsequently deployed based on the optimized task allocation strategy generated by the Geo-AI model. The simulation results show that using the proposed model significantly improves the allocation efficiency of rescue teams.
Conclusion: The Geo-AI–driven rescue model presented in this study offers a promising new avenue for enhancing the quality and efficiency of post-earthquake search and rescue efforts. Simulation results demonstrate that variables such as survival time under rubble, task completion time, proximity of rescuers to affected sites, and travel speed are critical to the effectiveness of rescue missions. Implementation of the intelligent model led to a 2.642-fold improvement in task allocation efficiency compared to traditional methods. These findings highlight the transformative potential of integrating artificial intelligence and spatial data systems into disaster response frameworks.
Keywords
Main Subjects
COPYRIGHTS
© 2025 The Author(s). This is an open-access article distributed under the terms and conditions of the Creative Attribution-NonCommercial 4.0 International (CC BY-NC 4.0)
)