Using AI To Create New Fragrances

Name: computer repairers
Brand: The Original PC Doctor
Rating: 4.8 (1226 reviews)

The market for fragrances is huge and diverse, with new formulas being released on quite a regular basis. By 2018, the global fragrance market had an estimated value of $70.70 billion. The market sees a 3.7% growth every year and plays an even bigger role compared to deodorants. As the market becomes more saturated, brands are finding it harder than ever before to come up with new ideas for fragrances that are sure to captivate the audience.

A new development in the artificial intelligence space might be the answer to these hurdles that brands are facing. IBM has developed an advanced AI system that is capable of analysing a wide range of data related to existing fragrances, raw ingredients, and other elements. The idea behind the technology is to help brands and individuals come up with new ideas and formulas for fragrances. We take a closer look at the technology in this post.

The Traditional Fragrance Development Process

Fragrances are exceptionally popular and used by a large percentage of the population. They allow you to feel more confident with a scent that you love. Many people also think of fragrances as something that attracts the right people. While commonly used, some still do not know how fragrances are made. Before we take a look at how AI is now becoming a formulating system for new fragrances, let’s consider the traditional way of making them.

The standard way of creating fragrances often involves complex processes. You would start by considering the type of fragrance that you wish to produce, then select ingredients that could give you the scent that you want to go for. A large number of fragrances on the market take advantage of natural resources, such as essential oils, fruit, and in some cases, they even use animal secretions. Coal, petrochemicals, alcohol, tar, and even wood are sometimes used to create a perfect fragrance.

The selection process, however, has been completely manual for a long time. This means people had to go out and do smell tests on a variety of ingredients, mix them together, and this often involved a lot of trial and error. In turn, the costs are often quite high, especially with fragrances that use some of the more exotic ingredients and extracts – as the creation of the fragrance often leads to large expenses and waste.

How AI Is Changing The Fragrance Creation Process

Artificial intelligence is no longer new to the modern day. Over the past couple of years, we have seen significant progress in the AI industry, along with adaptations in machine learning and related technologies. One of the latest inventions and trends involves the utilisation of artificial intelligence as a way of producing new formulas for fragrances.

IBM, the creators behind this new technology, explains that the ability to adopt an art for fragrance creation can take well over a decade to fully develop. The consumer’s main point that they consider during the purchase of a fragrance is the impression that it leaves. Without the right combination of creativity and science, it is difficult to create a scent that would leave a lasting impression.

IBM worked closely with Symrise on this particular project. Symrise is a global leader in the development of both fragrances and flavors. The company has been creating these formulas manually through the use of an expert team for several years already. With their expertise, IBM was able to utilise machine learning technology and feed the data to an artificial intelligence system that they developed.

There are four specific factors that AI focuses on:

Existing formulas have been created to produce fragrances.
The raw materials that are most often used to produce these fragrances.
Historical data about fragrance formulas that saw huge success.
Current trends within the fragrance industry, include the most popular fragrances on the market at the given point in time.

Data was provided by Symrise in order to facilitate the AI’s ability to analyse information and recognise patterns, as well as come up with new formulas.

This is not the first time that IBM dove into product creation technologies in its artificial intelligence department. The product composition project from IBM has been ongoing for some time and focuses on a diverse range of formulas and product options. Fragrances are one of the latest additions, and AI technology has already shown some promising results.

The process of AI work on fragrances. Image source: Cision News

Currently, the main focuses of the team behind the new AI technology focus on using machine learning as a way to learn and discover the following factors:

The ideal raw materials to be used in a formula and alternative raw materials that could be used as complementary elements or substitutes in a specific formula.
Identify the right dose of each raw ingredient used in a formula to achieve the ideal scent.
To help determine the expected response from human subjects exposed to a specific scent. The machine learning tech considers both gender appropriateness, as well as overall pleasantness.
To also provide a view of how the novelty of the specific scent would match with the commercial fragrances that are already available today.

IBM explains that one of their main objectives currently involves novelty when using AI to produce new fragrance formulas. When two fragrances are compared to each other, the closer the scent of each, the less novelty they have. This is why the AI was trained through machine learning to identify current scents that are on the market. The technique allows the AI to ensure the fragrance formulas it creates will not have a scent that is too similar to something already on the market.

For human subjects, including professional fragrance developers, creating this type of novelty in today’s market can be tough. There are thousands of different formulas available, so ensuring the mixture of ingredients used to create a new formula is not similar to any of these can be tricky.

When AI and machine learning have access to information about the scents and formulas of existing fragrances, they can process these at an incredibly fast rate. This way, companies and even individuals are able to come up with new formulas that will add that novelty element – by ensuring the scent is not something that the consumer has been exposed to ever before.

During the testing phase of the Philyra system, formulas were created by AI. These formulas were then assessed by professionals that form part of Symrise’s team, and a few tweaks were made in order to “perfect” each formula. Current plans involve introducing the system to master perfumers that form part of Symrise’s global system. The company also plans to use the system to create new fragrance formulas for a wide variety of home care and personal care products.

Conclusion

As technology and AI involve, people are finding new ways to use these inventions for business-related purposes. Philyra, a product composition technology, now allows people and businesses to easily formulate new fragrances by analysing a large database of information. While still in development, the technology already looks promising. More companies are also looking at this technology to help them bring to market a unique set of fragrances.

References

Written by The Original PC Doctor on 26/1/2023.

Embark on a journey of connection and joy! Share this page with your loved ones on your favourite digital platform. Click one of the icons below and let the magic of sharing begin!