Chemistry

Machine learning reveals complex odors are surprisingly easy to map

How the science connects

Machine learningOlfactionSensory system

AI Insight

Researchers have developed a machine learning approach to map complex odors in a systematic way, similar to how the Pantone system categorizes colors. This breakthrough addresses a long-standing challenge in olfactory science, as odors have historically been difficult to classify and compare objectively due to their chemical complexity. The machine learning model can predict and organize odor relationships more effectively than previously anticipated.


This development could revolutionize industries that rely on scent, including perfumery, food and beverage manufacturing, and potentially medical diagnostics where disease detection through odor is being explored. Creating a standardized odor map would enable better quality control, product development, and scientific communication about smells.


Understand the Science

Machine learning 184 articles Explore Concept → Olfaction Concept coming soon Sensory system Concept coming soon

If you want to describe a particular color, you could look to the Pantone color wheel to find its exact hue, saturation and brightness, and how it compares with other colors. But nothing like that has existed for complex odors.

Source: Complex odors prove easier to map than expected with machine learning