Evaluation and Analysis of a Novel Fugitive Road Dust Collector and Its Prediction Using Machine Learning Models

This research paper presents the conceptual design, development and performance evaluation of a compact fugitive vehicular dust collector to control and mitigate air pollution resulting from increased particulate matter on roadways. The developed dust collector is designed to work with two-wheeler automobiles, having a length 0.3 m, 0.15 m in diameter, and weight 2Kg, was experimentally tested on the gravel, asphalt, and concrete road at different speed of the vehicle. Results show distinguished dust collection efficiencies of 67.29% for gravel road, 65.65% for asphalt road, and 33.26% for concrete road, with dust collection increasing proportionally with speed of vehicle and maximum dust collection of 480 g at 80 km/h. The aerodynamic study verifies that dust jet length and mudguard dust area circumference expand as speed increases. The dust angle of attack increases from 220 at 20 km/h to 300 at 80 km/h, representing the direct relationship between speed of vehicle and dust dispersion pattern. The mathematical models were developed for gravel, asphalt and concrete road, reaching high correlations (R² > 98%), and machine learning approaches (Random Forest and XGBoost) further validated the accuracy of dust collection predictions. Relative assessment shows that gravel roads produce the more capturable dust, nearly 3% more than asphalt road and 35% more than concrete road. The dust collector is user-friendly, cost-effective (Rs. 7500 per unit, 83.94USD), and required very less maintenances with easy dust disposal. the mobile dust collector reduces fugitative particulate matter and minimises dust air pollution and provides a practical and scalable way to improve air quality, environmental sustainability and community health.