The smell of coffee could become part of a virtual catch-up, thanks to an experimental system that connects odor sensors to wearable scent dispensers over the internet.

Researchers at Ohio State University and Zhejiang Laboratory in China developed the automated system to identify smells using artificial intelligence and trigger their release near another person’s nose.

Described in Science Advances, the technology combines chemical sensing, machine learning, and wearable devices shaped as bracelets or necklaces.

In laboratory tests involving 14 healthy volunteers, participants identified individual odors with 73 percent accuracy and combinations with roughly 71 percent accuracy. The findings offer a step toward adding smell to digital experiences, although the system currently relies on preloaded scents.

Recreating smell presents a different challenge from reproducing images or sound. Cameras and microphones capture light and sound waves, while odors come from chemical compounds that must be detected and physically released for someone to experience them.

Previous approaches have included bulky desktop equipment, scratch-and-sniff cards, and virtual reality headsets. Some also required a person at the receiving end to release a scent manually.

The new system uses machine-learning algorithms to recognize patterns associated with volatile organic compounds, the molecules responsible for smells, and classify them into odor categories.

That information travels over the internet to a receiving device called Scentronic, which releases the corresponding scent close to the user’s nose. The process sends digital instructions between devices; the odor itself comes from supplies already loaded into the wearable.

The researchers tested five odors: beer, cake, coffee, orange, and red wine.

Source link