Foundation Models for Mapping Emission Sources and Acute Respiratory Infection (ARI) Hotspots

This study presents a cross-disciplinary approach combining foundation models, satellite imagery, and demographic health surveys to analyze the correlation between unregulated industrial activity and Acute Respiratory Infections (ARI) in South Asia. We detect brick kiln and factory chimneys using a sequential transformer chaining method and a multispectral Chimney Index. These detections are spatially joined with a geocoded dataset of ARI incidence developed from DHS surveys across India, Pakistan, and Bangladesh. Our findings reveal a statistically significant correlation between chimney density and ARI cases in children under five, underscoring the urgent need for regulatory and health interventions in high-emission zones.