Open Access

Review

Article ID: 3137

Enhancing urban disaster response through AI-driven data visualization for real-time decision support

by Kemisola Kasali

Public Safety and Emergency Management, Vol.1, No.1, 2025;

Urban areas face escalating threats from natural disasters due to climate change and rapid urbanization. This research explores how AI-driven data visualization enhances disaster response and supports real-time decision-making processes. The study proposes a reference architecture that integrates multiple data streams with adaptive visualization techniques, an advancement that improves situational awareness and coordination in emergency response environments. The research evaluates the Graph Attention Convolutional U-NET (GAC-UNET) model, which demonstrated high accuracy in flood detection tasks, achieving a 94% dice score and an 89% intersection over union (IoU). Case studies demonstrate practical benefits in real-world disaster situations, which include enhanced disaster impact prediction with greater precision, optimized resource allocation for maximum efficiency, and improved communication among diverse stakeholders in disaster response efforts. The findings reveal that AI-enabled data visualization significantly improves urban disaster response agility and accuracy, ultimately saving lives and reducing economic losses. The proposed framework adds operational capabilities to disaster management and offers improvements over traditional static dashboard systems. However, AI adoption in disaster management faces challenges such as data privacy concerns, security issues, ethical considerations in critical decision-making, and organizational resistance to new technology integration. This research emphasizes human-centered design principles and ethical AI governance frameworks for successful and responsible implementation. Future research should focus on generative AI models for scenario simulation, enhanced real-time predictive analytics capabilities, and community-driven platforms that improve collaboration and accelerate crisis decision-making processes.

(This article belongs to the Special Issue Technological Innovations in Emergency Management: Transforming Preparedness, Response, and Recovery)