Elucidating the Applications of Artificial Intelligence in the Context of Crisis Management

The escalating frequency, intensity, and complexity of natural and human-induced disasters worldwide have necessitated the adoption of advanced computational paradigms to enhance crisis management capabilities. Artificial intelligence has emerged as a transformative force across the disaster management cycle, offering unprecedented opportunities for prediction, situational awareness, resource allocation, and post-disaster recovery. This narrative review synthesizes the state of knowledge regarding AI applications in crisis management, drawing upon a curated selection of peer-reviewed literature published between ۲۰۲۰ and ۲۰۲۵. The review is structured according to the IMRAD framework and adopts the disaster management cycle as its organizing principle, examining AI contributions across the phases of preparedness, response, and recovery. A systematic search of Science Direct, Springer, IEEE, and SID.ir databases yielded ۱۲ high-impact articles that form the evidence base for this synthesis. The findings reveal six principal domains of AI application: hazard prediction and early warning systems (utilizing machine learning and deep learning for flood, earthquake, and wildfire forecasting); real-time situational awareness (leveraging natural language processing and computer vision on social media and satellite data); humanitarian logistics and resource optimization (employing reinforcement learning and optimization algorithms for evacuation and supply chain management); damage assessment (through semantic segmentation of remote sensing imagery); and public health crisis management. The review further identifies persistent challenges, including data scarcity and quality issues, model interpretability deficits, limited real-time deployment, ethical concerns regarding algorithmic bias and privacy, and the gap between academic AI solutions and operational practice. The article concludes by proposing a synthesized framework for integrating Al into crisis management systems and outlining priority directions for future research.

کلیدواژه ها:

Artificial Intelligence ، Crisis Management ، Disaster Management ، Machine Learning ، Deep Learning ، Humanitarian Logistics