The impact of road traffic on local air quality is a major public policy concern and has stimulated a substantial body of researchaimed at improving underlying vehicle and traffic management technologies and informing public policy action. Recent work hassought to use a variety of vehicle-based, person-based and infrastructure-based sensor systems to collect data on key aspects ofdriver and traffic behaviour, emissions, pollutant concentrations and exposure. The variety and pervasiveness of the sensor inputsavailable will increase significantly in the future as a result both of the increasingly widespread penetration of existingtechnologies (e.g., GPS based vehicle tracking, CANbus interfaces to on-board engine management system data) within thevehicle parc and the introduction of new technologies (such as e.g., UV sensing and nanotechnology based micro sensors). Aparticularly exciting direction for future development will be in the use of vehicles as platforms for outward facing environmentalsensor systems, allowing vehicles to operate as mobile environmental probes, providing radically improved capability for thedetection and monitoring of environmental pollutants and hazardous materials.However, these developments present new and formidable research challenges arising from the need to transmit,integrate, model and interpret vast quantities of highly diverse (spatially and temporally varying) sensor data. Our approach in thisproject is to address these challenges by novel combination and extension of state-of-the-art eScience, sensor, positioning andmodelling (data fusion, traffic, transport, emissions, dispersion) technologies. By so doing, we aim to develop the capability tomeasure, model and predict a wide range of environmental pollutants and hazards (both transport related and otherwise) using agrid of pervasive roadside and vehicle-mounted sensors. This work will be at the leading edge of eScience, stretching thecapabilities of the grid in a number of aspects of the processing of massive volumes of sensor data.