Olfactory intelligence: How AI is digitizing our sense of smell

Smell is connected to our emotions and memories in ways other senses aren’t. The smell of bread can transport us to childhood kitchens, while the scent of a loved one’s perfume offers comfort years later. It’s also our most ancient sense, evolved over millions of years to help our ancestors detect danger, find food, recognize kin and locate water.

Scent has long been a tool for both nature and humanity. In many ways, it is the language of nature itself. Flowers create beautiful scents attracting pollinators to help them reproduce. Dogs use their powerful noses to detect everything from smells as mundane as a ripe tomato to as critical as cancer cells.

Despite its importance, it’s one of the least understood senses, particularly in the digital world. For decades, computers have been able to see and hear, fundamentally changing how we communicate, work and understand our world. Yet computers are only just now learning how to utilize scent through the development of olfactory intelligence (OI).

OI could significantly improve healthcare, sustainability and the fabric of human connection.

Unlike vision’s three dimensions (red, green, blue) and sound’s one (frequency), smell operates across hundreds of receptor types, making it too complex for traditional science to map.

Digitizing scent couldn’t have happened without artificial intelligence (AI). With the help of graph neural networks, we built a map of smell, solving the mystery of which molecules created which scents. This made it possible for computers to then digitally record, analyze and reproduce those scents.

Each step is similar to how images are stored and recreated. You take a picture with a camera, download the image, sharpen or edit it, and then send it to a printer. Scent is captured with a laboratory device called a gas chromatography–mass spectrometry, or GC-MS, which is about the size of a dishwasher.

The GC-MS breaks down a scent into individual molecules, just like how a camera breaks down an image into pixels. The raw data is then converted into a scent encoding, which can be uploaded to the cloud.

While complex, olfactory intelligence is already being applied to a handful of industries from fragrance to healthcare and beyond.

The first test of olfactory intelligence is in fragrance, with the reproduction of the scent of a fresh-cut summer plum. The plum’s scent was captured and reproduced into a liquid scent that smelled identical. This opened the doors for the use of OI in the fragrance industry, changing the way fragrances are produced, while also serving as proof of concept.

With OI, fragrance development cycles that used to take years can be completed in days, allowing more time for the creative process, while also cutting costs that were often prohibitive for many brands.

With each fragrance developed, the technology gets smarter. The beautiful scents used to create perfumes, soaps and candles contain the same molecules as scents needed to better understand human health and happiness. The research continues to grow, and OI is now already proving impactful beyond just fragrance.

Oifactory intelligence is also addressing the costly problem of counterfeiting in the retail industry. Unlike existing authentication technology, OI can read the unique scent signatures of authentic products and identify counterfeits, recognizing subtle scent patterns, ignoring background scents and providing reliable answers.

Along with fragrance ingredients, olfactory intelligence has also been used to develop new insect repellents that are 10 times more potent than common active ingredient DEET. The World Health Organization (WHO) estimates that disease-carrying insects such as mosquitoes cause nearly a million deaths worldwide each year. These new repellents are directly addressing that issue.

These examples are just the beginning of where olfactory intelligence can take us.

Fragrances, counterfeits and insect repellents are just scratching the surface of what OI can do. The digitization of scent has numerous implications for human health and happiness.

Imagine a world where a handheld device has the same olfactory capabilities as a flower or a dog's nose, purifying air or detecting disease markers before symptoms appear. Where OI identifies viral outbreaks before pandemics and flags environmental health risks from pollution. The foundation for this is being built today.

This is no longer science fiction; these are real possibilities for our future. It’s safe to say that the future of olfaction is just beginning. With the help of olfactory intelligence, we will continue to decode the language of scent, not just building better technology but expanding human experience itself.