Fascination Technology Electronic Nose Detects Spoiled Food

Related Vendors

Taiwan Machine Tool & Accessory Builders' Association (TMBA)

SIMTOS (Seoul International Machine Tool Show)

In our section "Fascination Technology," we present impressive projects from research and development to engineers every week. Today: an electronic nose that reliably detects spoiled food and hidden allergens thanks to intelligent sensors and machine learning.

Many people rely on the smell test to decide whether slightly expired milk or leftovers from the previous day are still edible. But the human nose is not reliable enough: it does not detect all dangerous changes in food. The consequences are significant—in the USA alone, millions of people suffer from foodborne infections every year caused by spoiled or insufficiently cooked food.Researchers at UC Berkeley have therefore developed an "electronic nose" that works much more precisely. The device detects gases emitted by harmful bacteria and also identifies common allergens like walnuts and peanuts. This way, it can warn of health risks at an early stage.

One potential application is smart refrigerators: Equipped with appropriate sensors, they could actively notify users when food is about to spoil or is no longer safe to consume.

This is How the Artificial Nose Works

• 16 miniaturized gas sensors detect different gaseous compounds. 
• Each sensor has a specific coating and converts chemical reactions into electrical signals.
•  A machine learning model recognizes and interprets these signal profiles. From the signals, a characteristic "scent profile" is created—comparable to digital taste buds.
•  A machine learning model recognizes and interprets these signal profiles.
• It can distinguish between different foods such as strawberries, bananas, or nuts and also detects common allergens like walnuts, hazelnuts, cashews, and peanuts.
• Furthermore, it identifies the freshness of raw chicken, milk, and eggs—both when fresh and after 24 and 48 hours at room temperature.

High Sensitivity Proven in Tests

The results demonstrate high sensitivity: the system detects as little as 0.05 grams of walnut—about one hundredth of an average nut. However, there are still limitations in more complex environments. In the future, the device must also function reliably when different smells occur simultaneously, such as in mixed dishes or a filled refrigerator.

 "The idea behind it is that we can use the relative selectivity of gas sensors in combination with the pattern recognition capabilities of machine learning to determine which gas fingerprint corresponds to which food," said Bassil. "The result is a sensor chip that is far more sensitive and objective than any human nose could ever be."

Progress Thanks to Carbon Nanotubes

The concept of the electronic nose has existed since the 1980s. However, practical implementation long failed due to technical hurdles. Individual gas sensors—such as those in carbon monoxide detectors—can be manufactured relatively easily. Integrating many different sensors on a single chip, however, is more challenging.Carla Bassil from UC Berkeley achieved a significant breakthrough by using carbon nanotubes as conductive material. Unlike conventional metal oxides, they form extremely thin layers in the nanometer range. Their large surface area ensures high sensitivity—even at room temperature.This approach offers several advantages: the sensor does not need to be heated and can therefore work with a wider range of materials, including temperature-sensitive polymers. In addition, manufacturing becomes simplified. Instead of complex procedures, the so-called "drop casting" method is used, in which the materials are applied to the chip in a single step. 

The result is a scalable sensor design that efficiently combines different sensor materials, thereby laying the foundation for the practical use of electronic noses.