Informal transportation and health equity: Air pollution exposure in Latin American MTW microenvironments using neural networks

Introduction exposure to fine particles (PM 2.5 ) and Black Carbon (BC) is associated with adverse health effects, principally in low- and middle-income countries. This paper outlines the first empirical evidence in Latin America on traffic-related air pollution exposure within informal motorized three-wheeled vehicles (MTWs) and the health inequalities arising from this public transport mode. The objective of this research is to investigate the intersection of urban mobility, social determinants of health, and personal exposure to PM 2.5 and BC. Methods a cross-sectional study was conducted in Soledad, Colombia, a city characterized by rapid urban expansion and population growth with a high reliance on public buses and informal transport (38.1%). Data collection involved 21 O-D pairs using portable low-cost sensors and a microAeth, including GPS devices to capture real-time concentrations across three modes: buses, MTWs, and private cars. Inhaled doses were estimated by integrating pollutant concentrations with user demographic profiles and activity-specific inhalation rates. A Bayesian Neural Network Classifier (BNNC) was developed to model and predict high-risk exposure scenarios based on route, mode, and travel period. Results mean concentrations varied by mode, with buses and MTWs reporting the highest PM 2 . 5 levels of 33.0 and 27.5 μg/m 3 , respectively. Spatial analysis identified hotspots with peak MTW exposures of 959 μg/m 3 for PM 2 . 5 and 545 μg/m 3 for BC during rush hours. Results reveal that women may inhale doses PM 2.5 and BC up to 68 times higher than those inhaled by men across all transportation modes evaluated. A Bayesian Neural Network achieved 70% accuracy in predicting high-exposure scenarios. Conclusions These findings underscore that most commuters in the study area are exposed to levels exceeding WHO guidelines, especially those who use public and informal transport. The study highlights how urban mobility choices intersect with social factors, thus reinforcing structural socio-economic inequalities. These results demonstrate the pressing need for more equitable and sustainable urban transportation policies, and cleaner transportation strategies in rapidly developing cities.