Pervasive Air Quality Monitoring System Based On Low Cost Sensors

Air is one of the most important shared resources on our planet. Unfortunately, the quality of air has deteriorated significantly over the past years, especially for metropolitan cities in developing countries, such as Beijing and New Delhi. Among the various dimensions of air quality, particulate matter(PM) with diameters less than 2.5 micron, or P M2.5, has gained a lot of attention recently, partly because of its significant impact on our respiratory systems, and is the focus of this paper. Medical studies have shown that P M2.5can be easily absorbed by the lung, and high concentrations of PM2.5can lead to respiratory disease or even blood diseases.As a result, people are looking for better ways to monitor the quality of air in their immediate environment in order to take appropriate actions such as wearing masks or staying at home. While there are many smartphone applications that report publicly-available air quality data at the city or district-level, they cannot tell the actual air quality people breath-in, which is much more relevant and valuable. This is particularly important since we spend most of our time inside enclosed spaces such as homes and offices, where the air quality may deviate significantly from the outside. The government also need a dense deployment with accurate sensor readings to make strategies.In this paper, we present the design, implementation, and evaluation of Air Cloud –a novel client-cloud system for pervasive and personal air-quality monitoring at low cost.At the front-end, we create two types of Internet-connected particulate matter(PM2.5)monitors – AQM and mini AQM, with carefully designed mechanical structures for optimal air-flow. On the cloud-side, we create an air-quality analytics engine that learn and create models of air-quality based on a fusion of sensor data. This engine is used to calibrate AQMs and mini AQMs in real-time, and infer PM2.5concentrations. We evaluate Air Cloud using 5 months of data and 2 month of continuous deployment, and show that Air Cloud is able to achieve good accuracies at much lower cost than previous solutions.What’s more, we do a real dense deployment(about 200 AQM stations monitors)in Haidian district of Beijing, we then do a statistical analysis of the dense deployment sensor data and get some interesting and original findings, which may be very valuable in the future study area of PM2.5pollution tracing, locating the propagation path, etc. The result will help the government make more reasonable strategies to improve the air quality scientifically.