Credit: Shutterstock
A tiny chip developed at UC Berkeley is about to revolutionize food safety.
The University of California, Berkeley, has developed an “Electronic Nose” that can sniff out food spoilage and identify allergens with impressive accuracy.
A new proof-of-concept device created by @Cal_Engineer researchers could transform how food safety hazards are detected. The "electronic nose" identifies trace gases in the air, detecting signs of contamination before they're visible.https://t.co/p3WqlwzBga
— UC Berkeley (@UCBerkeley) June 17, 2026
The human nose is a remarkable organ, capable of distinguishing thousands of distinct odors. But it has limits that the food industry has long been aware of, and those limits carry consequences.
Food spoilage caused by bacteria produces volatile gases as they break down proteins and fats. These gases accumulate gradually, meaning that food can be in the early stages of bacterial contamination well before it smells bad. By the time human senses detect a problem, the food may already have entered unsafe territory.
Allergens present an equally severe challenge. Foods like peanuts, tree nuts, and shellfish are among the leading causes of life-threatening Anaphylactic reactions. Current detection methods require laboratory testing, which is time-consuming, expensive, and impractical for a consumer.
That gap between what the human nose can detect and what a device might detect is what the UC Berkeley team set out to close.
How the Electronic Nose Works
The device at the center of this research is officially called ML-SCENT (Machine Learning-Scalable Carbon Nanotube-based Electronic-nose Technology)
It contains 16 microscopic gas sensors, each constructed from carbon nanotubes, cylindrical structures of carbon atoms known for their extraordinary sensitivity to airborne molecules. Each sensor responds to a different set of chemical compounds. When the device is exposed to a food sample, each sensor generates an electrical signal whose strength and pattern reflect the specific volatile compounds present in the air around the food. Taken together, the 16 signals form a kind of chemical fingerprint.
That fingerprint is then analyzed by a Convolutional Neural Network, a widely used machine learning architecture for image recognition. The model has been trained to associate specific chemical fingerprint patterns with specific foods and conditions.
What It Can Detect
The research team tested the Electronic Nose across two broad categories of food safety concern.
In spoilage detection, the system detected chemical changes in common perishable foods over time and identified bacterial growth. Milk, eggs, and raw chicken were tested, and the device flagged spoilage at stages when visual and olfactory inspection would likely have marked the food safe.
In allergen detection, the system recognizes the chemical profiles of allergenic foods, including peanuts and tree nuts. The system could detect as little as 0.05 grams of walnuts.
Senior author Ali Javey, a professor of electrical engineering at UC Berkeley, has described the work as a demonstration of what is possible when novel nanomaterials and advanced machine learning are combined to address problems of practical importance.
A Device You Could Hold in Your Hand
One of the more striking aspects of the research is the team’s development of a portable electronic nose that connects to an iPhone.
This moves the technology out of the laboratory setting and into a format that could realistically be used by a grocery shopper evaluating a package of chicken, a parent checking a jar of peanut butter for traces of other allergens, or a restaurant manager running a spot check before dinner service begins.
For the roughly 33 million Americans living with diagnosed food allergies, a reliable, portable allergen detector would represent more than just convenience.
The researchers have suggested that the technology could eventually be embedded in smart appliances, such as refrigerators, that monitor their contents.
The Road Ahead
The research is published and peer-reviewed, and a portable prototype exists. But the path from laboratory prototype to widely available consumer product involves several additional steps. It requires further validation across a broader range of foods and conditions, regulatory assessment depending on the specific claims the product makes, manufacturing at scale, and integration into consumer-friendly hardware and software.
The researchers have not announced a commercial partnership or a timeline for market availability. The question now is how quickly the gap between that prototype and a product on a store shelf can be closed.
