Skip to main content

Lobelia Air: Low-cost air quality sensor calibration

The Lobelia Air project focuses on calibrating low-cost air quality sensors using AI techniques leveraging on EO derived AQ and meteorological data.


Business Category
Technical Category
Machine learning

What is the challenge that is being addressed?

Low-cost air quality sensors are a cheap and easily available alternative to expensive reference stations but show a degradation in accuracy over time. We demonstrate that the data enhanced with meteorological variables and background concentrations from CAMS can be automatically calibrated using AI.

What is the AI solution the project has implemented?

We implemented an extremely randomized tree trained on sensors in Sofia collocated with reference stations.

Who helped implement the AI solution?

This solution is implemented in the context of Lobelia Air, a winning project from the AI4Copernicus 3rd Open Call, by the Lobelia Earth company. The results of the project are available here.