Scentmatic, a Tokyo-based company delivering experiences that fully replicate scents and smells, in cooperation with Biotope Co. Ltd., plans to introduce its trademark Kaorium system to a limited number of Nose Shop‘s branches, a growing curator of perfumes and beauty products, nationwide in Japan.
Using AI to describe perfumes in the most artificially intelligent way possible.
Table of Contents
A Brief Introduction to Kaorium
Scentmatic is a company that aims to use the power of artificial intelligence for scent-related products and services. No, they don’t exactly sell colognes and perfumes using finely-calculated smart algorithms. Instead, the AI is used to “translate” these scents into words (relevant text), providing verbal profiles for each related product with the highest descriptive accuracy possible for the language chosen.
According to the company’s vision:
“The world is filled with different scents. But they are often not sensed properly, and thus the chance of fully experiencing them is often overlooked.”
The Kaorium system (from kaori/kaoru – scent/be fragrant) is the name of the specific AI used for the company services. As hinted, Kaorium doesn’t just specialize in colognes and perfumes. Just about any type of individual item or commercial product where scent or smell is of significant importance can be integrated into its articulation features.
Some of the product lines that Kaorium officially supports at the moment are:
- Chocolates – specifically different types of cacao products and its delectable combinations. In this manner, the Kaorium system may be used to put into words either the scent of the raw materials (for crafting reference), or actual product-offered chocolates (for special occasions, gifts, etc.)
- Sake – much like regular liquor, the scent of each type of sake is an important component of evaluation for its expected quality and impact. Kaorium for this application can be used to show potential customers what type of flavor combinations they might experience for the sake they’re planning to purchase.
- Artwork? – described by Scentmatic as “an additional avenue of recognition” for derived works and pieces. There are no specific tangible examples, but the word translation is supposed to enhance artistic entertainment and experience that uses all five major senses.
Collaboration with Nose Shop
Nose Shop is a rapidly growing specialty brand store in Japan that sells a very wide variety of perfumes catered for all types of customers. It was founded in 2017, and as of today, have six major branches across Japan. According to its (intended) business design, the brand has a unique focus on developing and introducing niche scents. This is then combined with rather extravagant and eye-popping marketing strategies to create buzz and drive interest into their latest masterpieces.
The partnership with Scentmatic is just but another of its wackier innovations. With the Kaorium system set to create a fully-detailed catalog of samples, the company will be offering its 5000+ customer data, in order to perfectly “simulate” the experience of its patrons into words that would fit the description of the chosen samples. At least 20 total samples have been confirmed, with the possibility of even more scents to be sampled using the advanced AI system in the near future.
As for the public demo, a pop-up sampling corner will open at its Shinjuku and Ginza branches. Opening date is on April 1st, and it will remain available until May 5th. Don’t expect the same touchscreen table you see here at the booth though. The one provided in this article is actually the concept design of the actual Kaorium system as a physical sampling device.
Nose Shop hopes that this collaborative demo can create “new avenues in choosing scents.” For Scentmagic, we suppose that this will also be a great opportunity to showcase the descriptive capabilities of Kaorium.
Welcome, Scent-Aficionado Robots?
As pretentious as they seem to sound when used by so-called connoisseurs, clear-cut descriptions of what otherwise qualities that can’t be (easily) put to words still provide an important evaluation point for certain products and services.
It won’t give you the exact idea. But at least you would have some idea of which type of wine to choose that would suit your taste. Or, it can show the occasion you are going to use them for, so long as these “expert” descriptions provide guidance.
That is why Kaorium as a concept doesn’t actually look as crazy or as impractical as it sounds. Using an adapting neural network AI that can shoot as many descriptions as it can might actually prove a lot more useful than either getting super generic descriptions, or overly ostentatious word profiles like “a tinge of this” or “a note of that.”
Besides, the idea of a machine learning bot listing out scent attributes for you may also turn out to be a very efficient experience time-wise. You get to know what you want a lot sooner, at least in theory.
The caveat, of course, is that a significantly more advanced iteration of the technology would be the more ideal “connoisseur” for these types of applications. The more experience it has, the better it would get at landing the descriptions accurately, instead of falling into the abyss of vagueness using near-universal adjectives.
True, there are limits to what words could do to specify the qualities of a particular item (try quantifying the word “transparent” to something you’re supposed to smell). But so long as it is suggestive enough, we can still probably use whatever something like Kaorium would provide as a basis to confirm our preferences.