Trendy young woman shows blue tongue and piercing against urban backdrop. (representational image)
Scientists have built a graphene-based device that can taste with near-human accuracy in a breakthrough that pushes artificial sensing closer to human ability.
The system uses machine learning to interpret chemical signals and classify flavors, even those it hasn’t seen before.
What sets this invention apart is its ability to work in moist conditions, a first for artificial gustatory systems.
This feature allows it to better simulate how real taste buds function inside the human mouth.
The device is built from layered graphene oxide within a nanofluidic structure. Unlike earlier attempts, it combines both sensing and computing into a single platform, making the system more integrated than traditional artificial tongues.
Graphene oxide, like pure graphene, reacts electrically when exposed to different chemicals.
The team trained the sensor using signals from 160 chemicals associated with common flavor profiles.
These signals were fed into a machine-learning algorithm that built a memory of how each flavor alters the material’s conductivity.
The learning approach closely mirrors how our brains process signals from taste buds. Humans were long believed to recognize five flavors such as sweet, salty, bitter, sour, and umami. In 2023, scientists added a sixth: ammonium chloride.
The artificial tongue focused on the first four. It identified previously learned tastes with about 98.5% accuracy. When introduced to 40 new flavors, its accuracy ranged between 75% and 90%.
Researchers also taught it to recognize more complex combinations, including those found in coffee and cola.
A conceptual diagram showing how a graphene-based artificial tongue mimics human taste by sensing chemical signals and processing them through a machine-learning system to identify sweet, salty, sour, and bitter flavors. Credit – Proceedings of the National Academy of Sciences.
Pure graphene was first isolated by Andre Geim and Kosta Novoslov in 2004, a feat that earned them the 2010 Nobel Prize in Physics. Its unique single-layer carbon atoms lattice structure offers outstanding electrical, mechanical, and chemical properties.
The new sensor leverages graphene oxide’s sensitivity to chemical changes. It detects slight variations in conductivity when exposed to flavor compounds, making it highly effective at pattern recognition when paired with machine learning.
“This system has the potential to one day restore taste perception to people who have lost that ability,” the authors noted. They added that loss of taste can result from stroke, viral infections, or degenerative diseases.
This innovation addresses a major limitation in earlier artificial tasting systems, the separation of sensing and processing. The current model’s unified structure allows for faster, more efficient interpretation of taste data.
However, the device remains a proof-of-concept. It is still too bulky and energy-intensive for consumer or medical use. The researchers say the next step will involve miniaturization and power optimization.
If successful, the sensor could find uses beyond healthcare, in food safety, quality control, and even robotics, wherever intelligent taste recognition is needed.
The study is published in the journal Proceedings of the National Academy of Sciences.
MasterCard
