Towards Large-Scale Drive-By Sensing With Multi-Purpose City Scanner Nodes

Drive-by sensing has the potential to provide hyperlocal data to study a number of urban phenomena and deliver actionable insights for public good. Yet drive-by deployments are often characterized by small fleets, huge costs and lack of flexibility to adapt to a city's multiple sensing needs. The City Scanner project aims to accelerate the development of driveby sensing by turning everyday vehicles into sensing nodes. In this paper we present Greta II, a solar-powered sensing platform that allows for low-cost continuous data collection and streaming of multiple phenomena. Thanks to a modular design it can be customized with off-the-shelf sensors and re-used across different deployments. Greta II can be easily built and deployed without disrupting a vehicle's normal operation. The platform has been validated during test deployments in two large American cities to collect air quality data. Drawing on our experience building and deploying Greta II we discuss challenges to be solved for enabling large-scale drive-by sensing.