Automated Machine Learning for Lower CO2 Emissions

Important results from the project

The project´s objective was met by publishing several scientific papers highlighting the environmental costs associated with machine learning experiments. These publications particularly highlighted how energy-intensive experiments can generate significant amounts of CO2 emissions, which contribute to climate change. By investigating and quantifying the energy consumption when using different machine learning algorithms, the project has shown the potential negative environmental effects.

Expected long term effects

The expected long-term effects include an increased awareness of the environmental consequences of energy-intensive technologies in research and industry. By highlighting carbon dioxide emissions linked to machine learning, the project hopes to inspire sustainable choices and methods. The goal is for energy efficiency to become a central factor in the development of new systems, which can lead to reduced emissions and increased responsibility within the technical sector.

Approach and implementation

The project was carried out through joint work during several visits by a foreign party to project manager. The work followed the schedule and the work proceeded without any unplanned or unexpected events.