AI Can Soon Tell Which Oil Your Chips Were Fried In

Ever bite into a snack and wonder if you're really getting the olive oil the label promises? A new study from Indian researchers shows how artificial intelligence, combined with a technology called Raman spectroscopy, can reliably identify which edible oil was used to cook processed foods like potato chips. Raman spectroscopy works by shining a laser on a sample and reading the unique "light fingerprint" that each molecule gives off. It's like reading the barcode of a substance — but the pattern is so complex that humans need AI to decode it.

In this paper, scientists tested five different edible oils in their pure form and also inside actual fried potato chips. They used several machine-learning methods, including decision trees (a rule-based approach that makes step-by-step choices) and K-means clustering (a way to group similar data points). For pure oils, the AI achieved 100% accuracy using just four key features from nearly 1,900 data points — a massive reduction of 99.44% of the data, with no loss of performance. That means the program isn't just memorizing; it's learning the physical signature of each oil.

On real chips, the job gets harder because paper and potato fibers muddy the signal. But when the researchers applied a "physics-informed" step — essentially breaking the signal down into its known chemical parts — accuracy improved to about 86%. That's not perfect, but it's a big step toward a practical tool. The model also reduced the data it needed to just four or five variables, making it compact enough to run on small, cheap devices.

Why should you care? Food fraud — like substituting expensive olive oil with cheaper palm or sunflower oil — is a global problem that hurts both your wallet and your trust. Today's testing is slow, expensive, and lab-bound. This research points to a future where a handheld scanner could check the authenticity of any packaged food at the factory, in a store, or even at customs. It's a promising example of what the authors call "Frugal AI": artificial intelligence that does more with far less computing power.