Researchers develop an AI-powered e-nose, offering a cost-effective and efficient way to detect and analyze oil spills.

Article published on: 11th November 2024

Credit: techxplore.com

In Summary:

A groundbreaking AI-powered e-nose, developed by researchers at the Skolkovo Institute of Science and Technology and international partners, can detect oil spills and determine their source based on the proportion of volatile organic compounds. This compact and portable device provides an efficient alternative to the expensive and bulky tandem gas chromatograph-mass spectrometers traditionally used. The e-nose, trained with an AI algorithm called a "random forest" model, was tested with oil samples from Kazakhstani fields and showed the capability to identify oil even after evaporation, making it valuable for environmental monitoring and compliance in the oil industry. The research, published in the Journal of Hazardous Materials, points to a future where industries can deploy the e-nose for practical field use and potentially train it to mimic human odor perception.

AI Explained: A random forest AI model is a type of ensemble learning method used for classification and regression tasks. It consists of multiple decision trees that work together to produce a single, more accurate prediction. Each tree in the forest is trained on a random subset of the data, and the final output is determined by averaging the results (for regression) or taking a majority vote (for classification) across all the trees. This approach helps improve the model's accuracy, reduces overfitting compared to a single decision tree, and increases robustness by combining the strengths of multiple models.

For the full article, visit the original post on: techexplore.com: AI-powered e-nose can detect oil spills efficiently

Credit: Journal of Hazardous Materials (2024) DOI: 10.1016/jhazmat.2024.135838