Real-Time Autonomous Systems for Tracking and Responding to Uncontrolled Fires in the Ambient Environment: A Review

TMLR Paper7170 Authors

Abstract: Fire-induced air pollution—originating from wildfires and industrial fires—poses a rising threat to public health and environmental systems. These episodic but increasingly frequent events release hazardous mixtures of particulate matter and gases, often overwhelming existing monitoring and response infrastructures. Traditional approaches to air quality sensing, health risk modelling, and emergency coordination are limited in spatial resolution, real-time responsiveness, and system integration. This literature review investigates how artificial intelligence (AI) and autonomous systems can address these limitations by enabling more adaptive, predictive, and interconnected fire pollution management strategies. Using a structured thematic synthesis, the review analyses 128 papers across four domains: (1) risks and impacts of fire-induced air pollution, (2) real-time autonomous systems for sensing, forecasting, and simulation, (3) AI-enhanced health risk modelling, and (4) governance and policy frameworks. Key findings reveal strong potential for UAV-based plume tracking, multi-agent learning systems, and data-driven health forecasting, but also highlight persistent gaps in regulatory readiness, system interoperability, and equity. The review argues for a coordinated, AI-centric framework to improve environmental sensing, health protection, and governance in fire-prone contexts.

Submission Type: Long submission (more than 12 pages of main content)

Assigned Action Editor: ~Marcello_Restelli1

Submission Number: 7170