Sharing smells online could become possible through Scentronic, a research prototype that combines an electronic nose, machine learning and a device capable of releasing fragrances at a distance. The system is designed to identify an aroma in one location and trigger a corresponding scent from equipment elsewhere.
Unlike photos, videos and voice messages, smell remains largely absent from digital communication. Scentronic addresses this gap by analysing the responses of gas sensors and classifying an odour with machine-learning algorithms. The result is then transmitted over the Internet to a receiver that releases a matching fragrance.
The original smell molecules do not travel through the network. Instead, the receiving device uses aromatic liquids stored in advance. For now, the prototype recognises five categories: coffee, beer, red wine, cake and orange. Expanding the selection would require additional system training and the availability of suitable fragrances in the receiving device.
The researchers stress that identifying an odour and recreating it accurately are different challenges. Detecting coffee at one end of a connection can cause a coffee fragrance to be released at the other, but this does not reproduce every characteristic of the specific coffee being detected. The system’s architecture is described in a study published in Science Advances.
In the reported tests, participants correctly identified individual aromas in about 73% of responses. For combinations of two smells, the success rate was close to 71%. The participants had first become familiar with the fragrances, and the researchers noted that the trials had limited control conditions. More rigorous testing will therefore be needed to determine how the approach performs in everyday settings.
The team also created receivers in bracelet, necklace and handheld formats. In addition, it tested an experimental cup that combines aroma release with the introduction of flavoured liquids. During a mixed-reality demonstration, selecting a virtual drink triggered a physical response from the cup, pointing to possible applications for adding smell and taste to experiences currently focused mainly on images and sound.
Scentronic remains a prototype with a small fragrance library, but it demonstrates one possible route towards transmitting selected aromas digitally.
Sources: Pplware.