Dust Alert Device: An Alternative for Real-time PM2.5 Monitoring to Support Disease Surveillance in Thailand with Performance Consistent Relative to Reference Measurements

Objectives: To assess the performance of the solar-powered, low-cost Dust Alert Device (DAD) for passive PM2.5 exposure screening and to evaluate its potential to strengthen air-quality surveillance in Thailand, where the limited spatial coverage of reference monitoring stations limits public health monitoring and risk communication. Methods: A cross-sectional study was conducted using six DAD units installed between January and December 2024 at the Department of Disease Control and five international checkpoints. Real-time DAD measurements were compared with data from the nearest Air4Thai reference stations. Linear regression analysis was used to examine the relationship between DAD and reference measurements. Mean absolute error (MAE) and mean absolute percentage error (MAPE) were calculated by comparing observed Air4Thai reference with predicted air quality values based on DAD measurements. Sensitivity, and specificity at multiple thresholds (25, 37.5, and 75 µg/m³ for PM2.5; 45 and 120 µg/m³ for PM₁₀) were also performed. Results: Hourly PM2.5 concentrations measured by DAD showed an MAE of 8.6 µg/m³ and an MAPE of 53.6%, and this reduced to 24.2% when PM2.5 exceeded 37.5 µg/m³ with a sensitivity of 79.6%. The performance of DAD was most consistent with Air4Thai under cool and dry environmental conditions. These findings suggest that DAD is a consistent surveillance-oriented tool that can practically strengthen PM2.5 monitoring and risk communication. Public Health Recommendations: Low-cost solar-powered sensors such as DAD can complement national monitoring networks by expanding spatial coverage, particularly in underserved or high-mobility locations such as border checkpoints. Integrating DAD data into routine surveillance and public communication platforms could improve early warning systems, support targeted health advisories, and enhance risk awareness among vulnerable populations. Further calibration methods and environmental adjustment models are recommended to improve accuracy across seasonal and climatic conditions.