From designer chairs to autonomous vehicles and identifying galaxies in deep space, it seems there’s nothing artificial intelligence (AI) can’t do. And now it’s being used to assess the quality of beer.
Researchers from the University of Melbourne’s Faculty of Veterinary and Agricultural Sciences and the Melbourne School of Engineering have developed an ‘electronic nose’, a low-cost device that uses machine learning to assess beer quality based on its aroma.
The e-nose consists of a small circuit board with a diameter of 92 millimetres, which can be placed face-down over a sample of beer in a beaker. Nine different gas sensors are incorporated into the device, measuring levels of components such as ethanol, methane, carbon monoxide, hydrogen, ammonia and carbon dioxide.
When coupled with a machine learning model, the presence of these gases can be used to predict the beer’s aroma and – due to the strong link between taste and smell – its quality.
“Our study shows that this technology can be used to assess quality in the field,” said lead author Associate Professor Sigfredo Fuentes.
“It’s a first for a genuinely portable device.”
To investigate the effectiveness of the e-nose, the researchers analysed 20 different commercial beer samples across a range of styles and fermentation types. Results were measured against a traditional sensory session conducted by 12 trained panellists, with positive results.
“With the e-nose, you just do a pour, and then you get all the data through the machine learning – you get the whole aroma profile of the beer with 97 per cent accuracy,” Fuentes said.
Future applications
While traditional techniques to evaluate aromas are often time consuming, requiring special training and costly equipment, the e-nose is low-cost and portable.
It allows brewers to undertake a rapid and reliable quality assessment on the production line, and eliminates the potential subjectivity associated with aroma testing sessions by brewers.
In the future, the researchers plan to refine the design to make the device smaller, and integrate rechargeable batteries and WiFi. The device could also have wider applications beyond beer, including building flavour profiles for sparkling wines, coffee and tea.
“The e-nose helps take a good deal of subjectivity out of things,” Fuentes said.
“Manufacturers will be able to definitively say what is in each flavour profile and put that information on the product labelling.”
