New ‘E-Nose’ Identifies Coffee with Up to 98% Accuracy

Researchers in Taiwan have developed an electronic nose (e-nose) that they say can correctly identify a coffee based on its aroma with up to 98% accuracy.

A team led by Chung-Hong Lee, a professor in the Department of Electrical Engineering at the National Kaohsiung University of Science and Technology (Taiwan) employed AI learning algorithms to the new e-nose.

Testing involved eight commercial metal semiconductor oxide sensors that detect specific gases associated with aroma. After being applied to roasted and brewed coffees of different varieties and origins, the sensors detected those gas levels, then sent the results to an AI algorithm for processing, according to a report in IEEE Spectrum, a sibling publication of the study publisher IEEE Xplore.

The outcome, according to the research team, was coffee identification accuracy ranging from 81-98%, depending on the type of coffee.

According to IEEE Spectrum — a publication of the international nonprofit institute of electronics engineers IEEE — the research team is using the new e-nose to develop a library of coffee aromas to improve the AI, while also catering to a broader diversity of coffee origins, processing methods and varieties.

“Beyond [my] personal enjoyment of coffee, the broader goal is to contribute to the preservation and understanding of aromas in the face of environmental changes, ensuring consistency in quality and flavor profiles across different crop years,” Lee told IEEE Spectrum.

The research team said it is currently seeking partners to continue the development of the e-nose, which may ultimately be used in any number of quality-control or verification applications throughout the coffee value chain.

Numerous other electronic noses and tongues have been explored as tools for use in the coffee, beverage and food industries over the years. Recent studies have also employed e-noses and e-tongues to asses differentiation in robusta coffee profiles, differences in roasted coffee profiles, and the effects of production methods on the sensory profiles of cold brew.

Kaohsiung University’s Lee is also developing an e-tongue for coffee tasting, an e-eye for assessing coffee cherry ripeness, and a tactile system for assessing green coffee’s moisture content.

Comments? Questions? News to share? Reach DCN’s editors here.