Real-Time Air Quality Monitoring and Pollution-Aware Route Optimization Using Vehicle-Mounted IoT Sensors.

Urban air pollution represents a critical public health crisis affecting billions of people worldwide, with the World Health Organization attributing approximately 4.2 million premature deaths annually to ambient air pollution exposure. Traditional air quality monitoring infrastructure relies on sparse networks of expensive stationary monitoring stations that fail to capture the fine-grained, route-specific pollution variations experienced by urban commuters. This paper presents the design, implementation and the experimental evaluation of the “Green Corridor” which is a comprehensive real-time air quality mon-itoring and pollution-aware route optimization system utilizing vehicle- mounted Internet of Things (IoT) sensors. The proposed system integrates an ESP32-based hardware sensing platform equipped with GP2Y1010AU0F optical particulate matter sen-sors, MQ- 135 multi-gas sensors, DHT11 temperature-humidity sensors and NEO-6M GPS modules for continuous environmental data acquisition during vehicle travel. The backend architecture employs a containerized microservices approach orchestrated through Docker Compose, featuring a high-throughput Go-based ingestion service achieving sub-50ms latency using the Fiber framework, a Python FastAPI routing service implementing weighted A* pathfinding with OSMnx graph operations, and Uber’s H3 hexagonal hierarchical spatial indexing system (Reso-lution 9, approximately 0.1 km2 per cell) for efficient geospatial data management. A novel “One Vehicle, One Vote” debouncing mechanism implemented through Redis Hash structures with time-to-live expiration prevents stationary vehicles from biasing aggregated pollution statistics. The routing algorithm imple-ments a parameterized cost function Cost(e) = T (e)(1 − β) + AQI(e)αβ, enabling users to dynamically balance travel time against pollution exposure through a continuous balance factor β ∈ [0, 1]. Experimental evaluation conducted over a two-month deployment period in Mumbai, India with three instrumented vehicles demonstrates the system’s capability to identify pollution hotspots with sub-second data processing latency, suggest cleaner alternative routes achieving 21-26% average AQI reduction with less than 13% travel time increase, and maintain reliable operation under varying network conditions. The frontend pro-vides an interactive React-based web interface with Leaflet.js map visualization featuring hexagonal pollution heatmaps, color-coded route comparisons, and quantitative exposure metrics. This work contributes a practical, scalable, and cost-effective solution for citizen-led participatory air quality monitoring and health-conscious navigation aligned with smart city sustainability objectives.