AI-based electronic tongue to advance food safety by detecting fruit juice and milk spoilage

Scientists in the US have developed an “electronic tongue” that leverages sensors, AI and machine learning to distinguish between similar liquids, such as milk with different water content, various types of soda and different coffee blends. The technology can also help the F&B industry enhance food safety by identifying signs of spoilage in fruit juices.

The sensor and its AI can broadly detect and classify various substances while collectively assessing their respective quality, authenticity and freshness, states the study.

“We’re trying to make an artificial tongue, but the process of how we experience different foods involves more than just the tongue,” says corresponding author Saptarshi Das, professor of engineering science and mechanics at Penn State, in the US, which led the research. 

“We have the tongue itself, consisting of taste receptors that interact with food species and send their information to the gustatory cortex — a biological neural network.”

To “artificially imitate” the gustatory cortex, the researchers developed a neural network, which is a machine learning algorithm that mimics how the human brain analyzes and understands information.

The findings, published in Nature, also indicate that the assessment has also provided the researchers with a view into how AI makes decisions, which could lead to better AI development and applications.

AI for food safety

The electronic tongue consists of a graphene-based ion-sensitive field-effect transistor or a conductive device that can detect chemical ions, linked to the artificial neural network the scientists have trained on various datasets. 

The electronic tongue comprises a graphene-based ion-sensitive field-effect transistor that can detect chemical ions (Image credit:Das Lab).The researchers provided the neural network with 20 specific parameters to assess, based on which the AI could accurately detect samples — including watered-down milks, different types of sodas, blends of coffee and multiple fruit juices “at several levels of freshness — and report on their content with greater than 80% accuracy in about a minute,” states the study.

The team previously investigated how the brain reacts to different tastes and mimicked this process by “integrating different 2D materials to develop a kind of blueprint as to how AI can process information more like a human being,” notes co-author Harikrishnan Ravichandran, a doctoral student in engineering science and mechanics advised by Das. 

The device was a graphene-based electronic sensor that can “taste” flavor profiles such as sweet and salty.

“Now, in this work, we’re considering several chemicals to see if the sensors can accurately detect them, and furthermore, whether they can detect minute differences between similar foods and discern instances of food safety concerns.”

Overcoming human limitations

During the product development, the team used game theory, a decision-making process that considers the choices of others to predict the outcome of a single participant, to assign values to the data under consideration. 

With these explanations, the researchers could reverse engineer an understanding of how the neural network weighed various components of the sample to make a final determination. This gave the team a glimpse into the neural network’s decision-making process, which they say has “remained largely opaque in the field of AI.”

This assessment could be compared to two people drinking milk, explain the researchers. “They can both identify that it is milk, but one person may think it is skim that has gone off while the other thinks it is 2% that is still fresh. The nuances of why are not easily explained even by the individual making the assessment.”

“We found that the network looked at more subtle characteristics in the data — things we, as humans, struggle to define properly,” states Das.

The artificial neural network and AI could accurately detect watered-down milks, juices and coffee blends at various levels of freshness.“In terms of the milk, the neural network can determine the varying water content of the milk and, in that context, determine if any indicators of degradation are meaningful enough to be considered a food safety issue.”

Versatility and cost factor

According to Das, the tongue’s capabilities are limited only by the data on which it is trained, meaning that while this study focused on food assessment, it could be applied to medical diagnostics and other industries, as well. 

He also explains that the sensors are not required to be “precisely identical” since machine learning algorithms can examine all information together and still produce the right answer. 

“This makes for a more practical and less expensive manufacturing process.”

“We figured out that we can live with imperfection. And that’s what nature is — it’s full of imperfections, but it can still make robust decisions, just like our electronic tongue,” concludes Das.

A Space Technology Graduate Research Opportunities grant from NASA supported the research.

Earlier this year, a Washington State University team in the US also developed an electronic tongue with strand-like sensory probes for early detection of microbial faults in white wines. The university also previously supported the tool’s role in detecting the difference of spiciness between samples of the same food.