Medindia LOGIN REGISTER
Medindia

AI-powered Tongue Capable of Detecting Food Safety Concerns

by Colleen Fleiss on Oct 12 2024 3:22 AM

AI-powered Tongue Capable of Detecting Food Safety Concerns
A new electronic tongue was found to detect subtle differences and signs of spoilage in liquids (milk, soda products, coffee blends, and fruit juices). The electronic tongue identified food safety concerns with greater precision. The results were published in the journal Nature. ()
Led by Saptarshi Das, the Ackley Professor of Engineering, the team designed the electronic tongue using a graphene-based ion-sensitive field-effect transistor associated with an artificial neural network.

A More Sensitive Taste: AI-Enhanced Electronic Tongue

This system was trained to detect the electrical properties of liquid samples. It identified their composition and freshness. AI defined its own parameters, which lead to even more accurate readings.

In tests, the electronic tongue achieved over 80% accuracy when using human-defined parameters and over 95% accuracy when the neural network assessed the data independently.

“Our goal was to mimic how human taste works, but with greater precision,” said Das. “By allowing the AI to assess the data holistically, it could identify more subtle changes that humans often struggle to define.”

Beyond food safety, electronic tongue has its role in medical diagnostics as well. The accuracy properties of the AI system make it a versatile tool for industries where quality, authenticity, and safety are critical.

The study received support from NASA through a Space Technology Graduate Research Opportunities grant.

Advertisement
Reference:
  1. Robust chemical analysis with graphene chemosensors and machine learning - (https://www.nature.com/articles/s41586-024-08003-w)
Source-Medindia


Advertisement