AI-powered wearables recreate five food smells sent over the internet

Researchers at Ohio State University and Zhejiang Laboratory built a system that uses machine learning to identify odors and send digital instructions over the internet to wearable scent dispensers called Scentronic. In tests with 14 volunteers, individual smells were identified with 73% accuracy and combinations with about 71%.

• The system covers five odors: beer, cake, coffee, orange and red wine
• Individual odors were identified with 73% accuracy, combinations with about 71%
• The system responds to a scent signal in 0.6 seconds
• Scentronic releases scents from preloaded odor channels