This AI system can predict air pollution before it actually happens

A group of computer scientists from Loughborough University in the UK has recently developed an artificial intelligence (AI) system that can predict the level of air pollution in a specific area. from days before the real pollution took place.

The only data to be used is the information obtained from the system of meteorological sensors installed in the city or any specific area. This data will then be automatically synthesized and analyzed to predict the potential pollution level with high accuracy.

This AI system is extremely effective in helping environmental experts to early assess air quality in densely populated areas, and promptly make early recommendations to people in polluted areas. Air often occurs, such as PM2.5 fine dust pollution. This type of dust has a very small diameter, only about 2.5 micrometers or even smaller. These dust particles can enter human body cells, destroying the defense and immune mechanisms from inside the cell causing a number of acute respiratory, heart and lung diseases. In areas contaminated with PM2.5 fine dust, visibility is greatly reduced and the sky is no longer as clear as usual. Early detection of this situation can help authorities take the necessary measures to minimize the situation, as well as people can actively protect their health.

Of course developing such a system is not simple. The researchers mainly focused on the sources of data obtained in China, as 145 out of 161 major Chinese cities regularly face problems with relatively serious air pollution. The history of air pollution in these cities is a rich source of data and is large enough to train highly efficient AI systems.

This AI system can predict air pollution before it actually happens Picture 1

In general, Loughborough University's AI system can help policymakers and scientists understand the causes of pollution patterns as well as explore ways to prevent or minimize the impact of they are based on the analysis of weather, season and environment data used in AI predictions.