The AI ​​program could help address the global microplastics challenge

The new AI program could address the global microplastic challenge

Credit: Hazardous materials magazine (2024). DOI: 10.1016/J.Jhazmat. 2024.136989

Monash researchers have developed a new AI program to help scientists in the global fight against environmental microplastics. The investigation is published in it Hazardous materials magazine.

In spite of the headlines in recent years, many policy scientists and formulators still do not know about the scale of the problem, including exactly what type of microplastics there are and where they are finishing.

The program developed by Monash uses sophisticated automatic learning algorithms to analyze thousands of samples in fractions of a second, a process that can take months for humans, to obtain a crucial understanding of where and how we need to act.

It is not as simple as placing the sample under a microscope, because the appearance by itself can be misleading.

For example, natural materials such as small pieces of sea shells can often resemble microplastics.

On the other hand, the new algorithm uses the chemical components that make up these materials to identify “firms” characteristics (complex numerical figures, many thousands of characters) that can accurately identify the types of known microplastics, using data from a process called infrared spectroscopy Fourier transformation (FTIR (FTIR).

Crucially, the program is the first in the world capable of analyzing a library of microplastic signatures, something desperately necessary by researchers who deal with the gigantic task of addressing the problem.

The advance was a pioneer by principal researcher Fithjof Herb, a Ph.D. from the University of Monash. Candidate and Supervisor Dr. Khay Fong, Professor of the School of Chemistry of Monash.

“We are addressing a significant bottleneck for progress when addressing the problem of microplastics,” said Herb. “Not only the process of analysis of samples consumes and slow, but so far, we have not been able to do it on a scale large enough to obtain an integral understanding of which microplastics we are dealing with, where they are and where they end.

“This is a very important first step to find ways to clean these harmful microplastics and find ways to prevent them from entering environmental river paths first.”

In addition to sea shells, other natural fibers commonly confused with microplastics include algae, animal skins or crustacean shells.

Herb said that the evolution of humans manufactured by humans also complicates things, with chemical components of microplastics that change constantly.

“Plastics constantly change, both in the way they are made and how they break down in the environment. Traditional tools fight to keep up with these changes,” he said. “But our tool offers a crucial advantage to scientists who need something that can adapt quickly, which is important to analyze the data that continues to evolve.

“We are really proud of what we have achieved here; it extends very well on conventional laptops, which reflects our approach to sustainability and accessibility, which we seek through small and efficient models.”

More information:
Fithjof Herb et al, automatic learning surpasses humans in microplastic characterization and reveals human labeling errors in FTIR data, Hazardous materials magazine (2024). DOI: 10.1016/J.Jhazmat. 2024.136989

Provided by Monash University


Citation: The IA program could help address the Global MicroPlastics Challenge (2025, February 11) recovered on February 18, 2025 from https://phys.org/news/2025-02-ai-tackle-global-microplastics. HTML

This document is subject to copyright. In addition to any fair treatment with the purpose of study or private research, you cannot reproduce any part without written permission. The content is provided only for information purposes.

#program #address #global #microplastics #challenge

Leave a Reply

Your email address will not be published. Required fields are marked *