Scentronic: Researchers develop AI system to transmit scents via internet

Researchers from Ohio State University and Zhejiang Laboratory have developed Scentronic, an experimental system designed to transmit odors digitally. The technology uses an electronic nose equipped with gas sensors to detect chemical compounds and machine learning to identify specific scents. Rather than sending physical molecules through the internet, the system transmits the scent's digital signature to a receiver.

The receiving device, which can be designed as a wearable bracelet, necklace, or handheld unit, uses a miniaturized atomizer to release pre-loaded aromatic liquids that recreate the detected smell. In laboratory tests involving 14 volunteers, the system achieved a 73% accuracy rate for single scents and 71% for scent combinations. The current prototype is limited to five specific aromas: coffee, cake, orange, beer, and red wine.

While still in the research prototype stage, potential applications include enhancing virtual and augmented reality, allowing online shoppers to test fragrances, and providing medical tools to assess olfactory function. Future development will focus on increasing scent variety, improving accuracy, and reducing the size and cost of the hardware.