Sign In

Communications of the ACM

ACM TechNews

Computer Scientists Develop AI System that Predicts Air Pollution Levels


View as: Print Mobile App Share: Send by email Share on reddit Share on StumbleUpon Share on Hacker News Share on Tweeter Share on Facebook
Air pollution.

A new artificial intelligence system developed at Loughborough University in the U.K. can predict air pollution levels up to two days in advance.

Credit: RT Magazine

Computer scientists at Loughborough University in the U.K. have developed an artificial intelligence (AI) system that can predict air pollution levels hours in advance.

The technology has the potential to provide information on the environmental factors that have significant impacts on air pollution levels.

The project is focused on predicting particulate matter less than 2.5 microns (10-6 m) in diameter, also known as PM2.5, a type of pollution often associated with reduced visibility in cities.

The AI system predicts PM2.5 levels in just a few hours, one to two days in advance. In addition, it interprets the various factors and data used for prediction, which could lead to a better understanding of variables that can impact PM2.5 levels.

From Loughborough University
View Full Article

 

Abstracts Copyright © 2020 SmithBucklin, Washington, DC, USA


 

No entries found