GeoAI-Based Air Pollution Exposure-Aware Route Optimization for School Commuting: A Comparative Study in Two Urban Environments

Air pollution is a major environmental risk factor, particularly for children, who experience repeated exposure during daily school commuting. Exposure-aware routing has been proposed as a strategy to reduce contact with environmental pollutants; however, limited evidence exists regarding how optimization opportunities vary across pollutants and urban environments. This study presents a GeoAI-based framework for pollutant-aware route optimization using OpenStreetMap street networks and air quality data from the Basque Government environmental monitoring network. Approximately 2000 simulated school commuting trajectories were generated across Donostia-San Sebastián and Bilbao’s metropolitan area. For each origin–destination pair, the shortest-path route was compared with alternative routes optimized for PM10 and NOx exposure. The results revealed substantial differences between pollutants and cities. In Donostia-San Sebastián, NOx optimization produced the greatest benefits, with more than 10% of routes achieving exposure reductions above 5% and a maximum reduction of 54.9%. In contrast, Bilbao exhibited the highest optimization potential for PM10, with a maximum reduction of 32.8% and 8.2% of routes achieving reductions greater than 5%. Meaningful reductions were generally achieved with moderate increases in traveled distance. These findings suggest that exposure-aware routing may benefit a subset of school commuting trajectories and that optimization potential strongly depends on the local environmental conditions.