AI-Powered Wearable Device Could One Day Transmit Smells Over the Internet

Our sense of smell is a vital way of experiencing the world, often connected to memories and emotions. Despite its importance, recreating smell digitally remains a challenge. Unlike vision and sound, which are now easily reproduced in our homes, smell can't yet be digitally replicated, meaning we can't share experiences like enjoying a cooking show or exploring a forest floor virtually.

The complexity of olfaction lies in sensing chemical compounds, a task that presents significant hurdles in capturing, transmitting, and recreating smells.

Researchers at Ohio State University and Zhejiang Laboratory in China have taken a significant step forward in this area. They've developed a system that could transmit smells over the internet using a wearable device. Their findings, published in Science Advances, describe a system that employs artificial intelligence to identify the chemical compounds responsible for specific smells. The device consists of wearable accessories, such as bracelets or necklaces, which release the transmitted scent near the receiver's nose.

In their experiments, the team exposed 14 healthy volunteers to five distinct odors (beer, cake, coffee, orange, and red wine) individually or in combinations. The volunteers were asked to identify these smells. The system successfully identified individual smells with a 73% accuracy rate and correctly identified smell combinations 71% of the time. The device responded to scent signals in just 0.6 seconds, a speed comparable to a heartbeat. However, the system required more time to accurately classify complex food odors.

Building on their initial system, the researchers also developed a version capable of dispensing flavored liquids infused with scents, aiming to recreate both smells and tastes transmitted via the internet. They envision future applications for this technology in virtual social interactions, online shopping, and healthcare, including psychophysical testing for sensory loss treatment.

However, the team acknowledges the system's limitations, such as the lack of precision due to the use of preloaded odor channels, and plans to test the system in real-world settings and further expand its capabilities.