Smell Digitization Opening Up New Businesses

Vision and hearing are being digitized, and video content and music content are becoming big businesses. On the other hand, among the five human senses, the sense of smell is the least digitized. The smell related market is said to be worth over 200 trillion yen worldwide, but digitization businesses now account for a small part of that. We are conducting joint research by combining the world's first technology for quantifying smells with NTT DATA's quantum computing technology. We introduce our efforts, results, and future prospects.

We are planning to hold a demonstration at the Osaka Kansai Expo, where you can experiencethe future of smell digitization.

1. Business Opportunities with Smell

The largest potential market for businesses related to smell exceeds 200 trillion yen worldwide. The market for fragrances and other products associated with smell is tens of trillions of yen. Smell also has business potential in pest control and the diagnosis of specific diseases, such as cancer. The market size for digital scent technology is only around 300 billion yen.

Digitizing smells is more complex than the visual and auditory senses, making it difficult for businesses to expand. However, with the quantum computing technology that NTT DATA has cultivated, it is possible to overcome this complexity hurdle and expand the business.

2. Why Smells Are So Difficult to Digitize?

Originally, artisans with years of experience and training, called perfumers, used their sense of smell to mix fragrances. The five senses of human beings differ from person to person. Therefore, not only ordinary people but also trained perfumers perceive the same smell differently.

What is the sense of smell? Vision is the function of detecting light frequencies and perceiving them as colors. Hearing is the function of detecting the frequency of air and perceiving it as sound. And what is the sense of smell?

The sense of smell detects the shape of volatile small organic compounds called smell molecules and perceives them as a smell. Different substances with similar molecular shapes are perceived as similar smells. Similar substances with different molecular shapes are perceived as a completely different smell. With this more complex nature than other senses, smell is difficult to digitize.

There are two main ways to digitize smells. One is e-nose technology using sensors. The other is digital technology using brain wave readings. However, both technologies have major drawbacks. E-nose technology uses sensors to detect molecules themselves. Because it recognizes individual molecules rather than the smell, it requires as many sensors as there are substances. It is estimated that there are hundreds of thousands of olfactory molecules, so it is not possible to cover all smells. In digital technology using brain wave readings, the brain connects the odor molecules with the emotion of the smell, rather than the smell itself. For example, it can be used to develop relaxing aromas, but the development of scents itself must rely on years of experience.

3. Komi-Hakko, a Company That Quantifies Smell

Komi-Hakko is the world's first company that has the technology to quantify smells in a completely different way from the conventional smell digitization technology. Komi-Hakko is a venture company from The University of Osaka with a mission to implement this smell quantification technology in society.

There are cells called olfactory receptors that detect smell. Komi-Hakko has succeeded in producing cells that mimic these olfactory receptors and can quantify the smell responses that humans detect. In fact, there are various examples of digitization technology initiatives focusing on olfactory receptors. Then, why is the technology of Komi-Hakko the first in the world?

Other olfactory receptor digitization techniques remain limited to a subset of olfactory receptors and still cannot take a temporal response to see if a response has occurred. Therefore, it is not possible to achieve accuracy in smell digitization. Komi-hakko, on the other hand, has the only technology in the world that can quantify the temporal response of all olfactory receptors from the beginning to the end of the response.

When one wants to develop a specific scent, this quantification technology is used to quantify that scent. By combining multiple fragrances to match the data, it is possible to create a smell that matches the target. This enables reproducible smell development independent of the perfumers’ experience. Aroma Fermentation has created a smell database by quantifying more than 2,000 types of flavors and fragrances. Based on this database, we can conduct various businesses related to smell development. Komi-hakko has created a smell database by quantifying more than 2,000 fragrances. Based on this smell database, businesses related to the development of various smells can be conducted.

The database will expand to 8,000 fragrances in the future, but there will be a problem. There are about 40 types of olfactory receptors. 8,000 fragrances must be selected and combined for about 400 responses of the target. The number of combinations is enormous, and the sense of speed that is the advantage of digitalization will be lost.

4. Smell Reconstruction Using Quantum Computing Technology

NTT DATA is conducting research and development on quantum computing technology. A quantum computer is a computer that uses quantum mechanical phenomena and is expected to be much faster than conventional electronic computers. There are two main types of quantum computers: quantum gate and quantum annealing. Quantum gate is a computer for general-purpose computing, and basic research is actively underway. Quantum annealing is a computer for combinatorial optimization and is being applied to real businesses. This quantum annealing method is expected to solve the big problem of smell reconstruction.

Combinatorial optimization is the academic field of finding the combination that minimizes and maximizes a certain index. For a problem with 1 million possible combinations, it guarantees what kind of combination to search to find the optimal combination. This combinatorial optimization is attracting attention from business because the computational time has been greatly reduced by the advanced performance of computers. However, it is still not uncommon for such calculations to take several dozen hours, making computation time a major issue.

A quantum computer based on quantum annealing can solve the optimal solution of this combinatorial optimization problem in an instant by using the superposition phenomenon of quantum mechanics.

NTT DATA and Komi-hakko conducted joint research, and using NTT's quantum annealing LASOLV, we were able to perform smell reconstruction in a short time, which used to take a long time to calculate. This eliminates barriers to smell reconstruction and accelerates business applications.

5. Business through Smell Reconstruction

There are two major businesses related to smell reconstitution: the use of smell reconstruction itself and the deodorization field. In conventional smell development, for example, the smell of vanilla beans was created by extracting the components contained in vanilla beans. However, by using this smell reconstruction technique, it is possible to create the same smell by composing substances different from those really contained.

By including conditions such as using only environmentally friendly substances, composing with as few kinds of substances as possible, and using inexpensive substances, it is possible to develop fragrances that have contributed to many social problems.

Furthermore, since smells are digitized, it is possible to transfer smell composition data over the Internet to reproduce the smell at a distance. In recent years, improvements in Internet technology have made it possible to share visual and auditory information remotely, but it is also possible to share olfactory information, making it possible to share experiences more realistically.

Smell also plays an important role in eating. The sense of taste is composed of five flavors (sweet, umami, bitter, salty, sour), and the sense of smell is added to express rich tastes. Everyone has probably done an experiment at least once in which the taste of rice is not felt when eating with the nose covered, so the degree to which smell relates to taste is that great. Conversely, as long as the smell is strong, even people with a weak sense of taste can sense rich tastes. By adding flavorings that amplify the taste of insipid hospital food, it is possible to enjoy a pleasant meal without affecting health.

This smell quantification technology proposes a completely new method of deodorization. Conventional methods of deodorization have involved either adsorbing smell molecules with activated carbon or applying a stronger smell to prevent the smell from being felt. The response of smell molecules to olfactory receptors is known as the allosteric effect, a phenomenon in which multiple molecules react simultaneously to intensify or weaken the smell. Using this phenomenon, it is possible to develop a pinpoint deodorant that attenuates or neutralizes a specific smell. By loading this deodorant into a mask, factories and other facilities can block only a specific smell and provide comfort.

6. Future Prospect of Smell Digitization

With further advances in smell reconstruction technology, for example, it may be possible to develop repellents that only mosquitoes can sense unpleasant smell by recreating the olfactory receptors of mosquitoes, thereby reducing diseases that spread through mosquitoes around the world. In addition, by developing tools that can detect a unique smell that doctors can sense when they are sick, such a tool may make it possible to easily perform diagnoses like professional doctors, building a healthier society.

Smell digitization technology, which has been slow to make progress until now, could contribute not only to entertainment but also to various social issues such as healthcare and the natural environment.

The scent reconstruction is scheduled to be exhibited at the Osaka Expo, which is currently underway, so you can feel the future of scent digitization.

Kaiki Ii

Innovation Center, Innovation Technology Department, Technology and Innovation General Headquarters, NTT DATA Group

Engaged in R&D and system development to utilize cutting-edge technologies such as mathematical optimization and quantum computers in actual business, and is responsible for developing applications for mathematical optimization and evaluating and verifying advanced technologies.

Takashi Yazane

Innovation Center, Innovation Technology Department, Technology and Innovation General Headquarters, NTT DATA Group

Engaged in R&D and system development for applications in the field of operations research, such as mathematical optimization and simulation techniques, etc. Since 2018, he has been engaged in R&D of quantum computing and verification support services for customers through the Quantum Computing / Next Generation Architecture Lab