Breathing in the digital age: How AI can revolutionize air quality monitoring

In 2023, air pollution contributed to more than 7 million deaths worldwide. But in the era of artificial intelligence (AI), smart and efficient air quality monitoring systems could be developed to detect air pollutants more quickly and accurately.

Ground-based air quality monitoring systems provide data on the level of air pollutants present in the troposphere – the densest layer of Earth’s atmosphere, which extends up to about 8 miles from the surface. But machine learning (ML) algorithms can be paired with current air quality monitoring systems to track changes in the atmosphere and provide early warning alerts about air pollution hazards.

Integrating AI with existing air quality monitoring systems in this way can create improvements in four key areas:

AI and ML models can process large volumes of data, such as those currently generated by air quality sensors. This can be done at a faster speed to quickly detect changes in the levels of pollutants in the air, allowing for faster action by governments trying to improve air quality.

AI models can use sensor data to make air quality monitoring information more accessible to a larger audience. Automating this process with AI reduces the cost of human intervention in data analysis.

AI algorithms can enhance data collection and analysis of air pollutants by ensuring users receive more precise information. Recent research has shown that the accuracy of air quality forecasting can be improved by ML models.

AI models can provide data-driven insights to help government bodies and businesses to make informed decisions faster, protecting people from hazardous levels of air pollution.

Many nations have already started to realise these benefits as they integrate AI into their air quality monitoring systems. It is helping them to collect more real-time data that can be analyzed and used to make decisions to protect people from the effects of air pollution.

South African scientists have developed Ai_r, for example. This relatively cost-effective system monitors air quality and makes predictions about future pollution hotspots. An automated air quality forecasting system named AI-Air has also been developed by Chinese researchers, which aims to improve the forecasting of pollutant concentrations.

AirQo is another example of a current system in this area that’s being used in Africa. It aims to use AI to inform decision-making about ways to tackle air pollution health challenges across more than 16 African cities. The platform does this by combining low-cost sensors, AI algorithms and user-friendly interfaces to analyse air quality insights.

Satellite-based air quality monitoring is another developing area. A team of researchers from China and Japan has developed AIRTrans, for example. This AI-powered tool has successfully used satellites to capture information about aerosol concentrations and their size, making it an effective solution for pollutant monitoring and early warning systems. It can help to predict pollution trends for a particular city by analysing previous datasets.

In Korea, researchers have used a combination of AI algorithms to forecast air pollutants and develop air quality monitoring and early warning systems, while similar AI-driven pollution forecasts saw forecast accuracy rise to 92% within 18 months when used in China.

Despite this progress, there are still challenges to integrating AI into air quality monitoring systems. AI models need large volumes of accurate datasets to train effectively, and the availability of such data can become a constraint. Setting up an AI-based air quality monitoring system is also very costly because they require data centre resources and large amounts of electricity.

In addition, there is a lack of availability of skilled personnel for the development of ML algorithms and sensor hardware maintenance. Integrating AI-driven air quality monitoring systems into existing infrastructure can be expensive and complex too.

These problems must be addressed for AI-driven air quality monitoring systems to be effective, accurate and affordable.

By providing real-time and predictive analysis, AI is already revolutionizing air quality monitoring and forecasting efforts around the world, which could help to achieve sustainable development goals. Future innovations could include predictive models for air quality monitoring and more use of IoT sensors with existing infrastructure.

Further, AI-powered drones could help detect air pollutants in hard-to-access or remote areas and the data they collect could be analyzed using AI algorithms. Smart cities and IoT could also enable the deployment of networks of low-cost air quality sensors. These sensors would provide continuous, real-time data on a city's air pollution levels.

AI is revolutionizing air quality monitoring systems by enabling real-time, high-resolution data analysis. Through integration with Internet of Things (IoT) and big data, air quality monitoring systems can become more efficient. This advancement in air quality monitoring systems would allow governments, institutions and environmental agencies to take timely decisions and improve public health.