The Real-Time Data Challenge Managing Thousands of Air Pollution Sensors

Featured article by Yann Boquillod, founder, AirVisual

The quality of the air we breathe is one of the leading environmental health hazards facing our global population. Last year the World Health Organization (WHO) released a global air quality model reaching the alarming conclusion that 92% of the world’s population is currently breathing air that’s polluted enough to represent a health risk. Additional WHO data estimates that 3 million deaths annually are attributable to outdoor air pollution, with two-thirds of these deaths occurring in the heavily polluted South East Asia and Western Pacific regions.

By recognizing the sources of air pollution and gaining insights into shifts in air quality, measures can be taken to mitigate the pollution’s effects and prevent exposure. Improved data gathering not only illuminates the challenge of air pollution’s impact on global health, it also contributes to the solution. We founded AirVisual to serve this purpose, providing the public with historical and real-time air quality information, as well as high accuracy forecasts.

To accomplish this, AirVisual gathers air quality data from a worldwide network of sensors in over 8,000 cities worldwide. This network comprises both data from public, government monitoring stations as well as a growing network of crowdsourced monitors from the AirVisual community (whereby community members deploy AirVisual Node air quality monitors as a public, outdoor station). In this way, the technology strives to make air quality data easily and freely accessible to people everywhere, while also empowering individuals to broadcast their community’s local air quality and contribute to the global air pollution map.

[Real-time air quality data around the world.]

Equipped with a dynamic database of sensor readings from around the world, AirVisual is able to utilize this rich resource to perform complex air quality and data modeling. AirVisual was the first to begin generating an air quality forecast, enabling people to prepare and protect themselves against polluted outdoor conditions. More recently, we created the world’s first 3D, real-time interactive map to display the Earth’s health. Combining sensor information with satellite data, AirVisual Earth overlays weather patterns with PM2.5 pollution to provide a uniquely engaging representation of how pollution is affecting the state of our planet, as well as ourselves.

[https://airvisual.com/earth.]

 Putting AirVisual’s data into action, individuals and customers are able to make informed, real-time decisions to mitigate exposure to dangerous pollution. The air quality data gathered by the sensor network can be freely accessed by the public through the AirVisual website or free mobile air quality app. Enterprises can also incorporate the data within their own services, utilizing AirVisual Node sensors to monitor air quality at their facilities. In this way, business offices, schools, government agencies and other locations are able to operate with an awareness of their indoor and outdoor air quality, improving the effectiveness of ventilation and air purification procedures, and ultimately, health outcomes.

Central to all this is providing reliable, accurate data. To achieve the technical feat of building and maintaining a database able to handle the real-time nature and high availability needs of the service, we decided to run MongoDB on Amazon Web Services. We also opted to make use of a hosted MongoDB database service in order to maximize the output from our resources for developing the service for our community.

For that we chose mLab’s Database-as-a-Service platform, which has been a great fit for us. Critically, we needed to ensure low latency and a great experience for our Asia-based community members (who can be exposed to some of the worst air pollution on the planet); the platform allows us to create database deployments in the AWS Japan data center. The platform’s database monitoring capabilities also help ensure that our database operations and queries are optimized, as well as the peace of mind in verifying, with a quick check, that everything is running smoothly. As a result, we’ve been able to scale our databases quickly as we continuously add more global data and deploy more crowdsourced sensors to bring awareness to new communities across the world.

The air quality data AirVisual measures and provides is aimed to give individuals and organizations a better understanding of the dangers of air pollution in their areas, and to take smart steps in response. As real-time information and reliable forecasts increase awareness of local air quality, we hope that the impact of this knowledge will be to improve the health and well being of individuals worldwide, increase the successfulness of anti-pollution measures, and save lives.

Yann Boquillod is the founder of AirVisual, a crowdsourced social enterprise offering real-time air quality monitoring, forecasting, and visualization services to individuals and enterprises.