Monitoring environmental features, such as air pollution, carbon dioxide emissions, noise, and heat, gives cities key datadriven insights to advise sustainable policies and city design. However, given the high variability of the... more
Monitoring environmental features, such as air pollution, carbon dioxide emissions, noise, and heat, gives cities key datadriven insights to advise sustainable policies and city design. However, given the high variability of the environmental data, achieving good spatio-temporal resolution and coverage remains a major challenge. Even in well-monitored cities, such as Amsterdam, environmental sensors are usually placed in very few fixed locations, implying limited spatial coverage and an inability to adapt to changes in the urban environment. As cities evolve, they experience shifts in pollution sources, and fixed sensors might not adequately capture these changes without a costly and time-consuming reconfiguration process. To monitor the environmental qualities of Amsterdam's roads, we present a ''drive-by'' sensing solution for a structured network of vehicles, meaning that sensors are designed to be deployed on buses and tramways, the trajectories and schedules of which are known. We propose a deployment strategy that combines the available fleets to reach optimal spatio-temporal coverage for different environmental features. For example, by optimizing the deployment of sensors on public transit vehicles, our proposal significantly enhances the monitoring of pollution-sensitive areas in Amsterdam. Depending on the desired spatiotemporal granularity and noting that one vehicle only hosts one sensor, the required number of sensors to be deployed on the structured network varies between 43 and 142, with the latter achieving the finest possible resolution.
