Kirin Holdings’ AI and molecular simulation collab: “Double cleansing is the key to beautiful skin”

24 Nov 2022 --- A Japan-based research collaboration between Kirin Holdings, FANCL and Keio University has resulted in an enhanced cleansing agent developer utilizing artificial intelligence (AI) and molecular simulation technologies. This can be used when developing a cleanser with a “high makeup removal function.” The cleansing and washing process is expected to reduce cleansing time, save water and thus, lower environmental impacts.

“Since cleansing alone removes dirt and grime and refreshes the skin, some may wonder if they can get by without a face wash. Some may feel that they do not need to wash their face. However, since cleansers target different types of dirt (oil-based and water-based), cleansing with a double face wash is essential for skin care,” Hajime Ohtsuka, office of the president, general manager of public relations at FANCL, tells PersonalCareInsights.

“On days when only light makeup or sunscreen is used, it is easy to wash your face without cleansing, but no matter how light the makeup is, it can only be removed by cleansing. Sunscreen, in particular, is applied to the face so thoroughly that dirt is embedded in the pores. The key to beautiful skin is to thoroughly double-cleanse your face to remove dirt and grime.”

Tech takeover
Based on experiments by FANCL, it was found that the developer could quickly and efficiently predict the performance of cleansing agents with “high accuracy.” The developer was built with a prediction system to show which combined ingredients are better for cleansing makeup. 

“Cosmetics development requires a great deal of time, effort, and experience on the part of researchers to find combinations that ensure the desired functions and safety from among the myriad of raw materials that exist,” says Ohtsuka.Combining EH with PPG-9 diglyceryl ether and cyclohexylglycerin led to high cleansing rates. Credit: Kirin Holdings.

“It has been suggested that the use of AI technology has the potential to significantly increase the speed of finding and developing optimal formulations.”

The machine learning technology by Kirin Holdings analyzes and learns from data, and then finds patterns within it. The computer can then predict the makeup removal performance of an unknown cleansing agent by learning the data of its raw materials, says the company.

On the other hand, Keio University’s molecular simulation technology analyzes molecular phenomena by reproducing the movement of atoms and molecules in virtual space on a computer.

Best performing ingredient
In the study, 100,000 calculations were performed on a computer using Kirin Holdings’ machine learning technology on the developer system.

FANCL’s eicosaglycerol hexacaprylate (EH) ingredient had the highest makeup removal performance among the surfactants. “A cleansing ingredient classified as a nonionic surfactant. It is a raw material and has been used in products such as mild cleansing oil,” notes Kirin Holdings. 

Additionally, combining EH with ingredients like PPG-9 diglyceryl ether and cyclohexylglycerin – moisturizing agents – led to high cleansing rates.

As a result of the highest predicted cleansing rates, a prototype cleansing agent containing the ingredients mentioned above was created and used to remove makeup. Kirin Holdings reports that the predicted and actual cleansing (on skin with makeup and dirt like sebum and mud) rates “generally matched.”

Cleansing is not the same as facewash
Ohtsuka tells us that there is a difference between cleansing and face wash and that double cleansing is the key to beautiful skin. 

“Cleansing is to remove greasy dirt such as makeup, keratin plugs, blackheads and other dirt clogged in pores,” he clarifies.

“In contrast, face washing removes dead skin cells, excess sebum, sweat, and dust on the face. It may be easier to understand that cleansing mainly removes ‘oily’ dirt and stubborn pores that are difficult to remove, while face washing mainly cleanses ‘water-based’ dirt.”

Commercial plans
FANCL plans to continue to add information obtained from molecular simulations to machine learning to improve its prediction accuracy. The developer was built with a prediction system to show which combined ingredients are better for cleansing makeup. Credit: Kirin Holdings.

The beauty corporation will also consider commercializing cleansing agents with high makeup removal functions. “In addition, the research will continue with a view to horizontally expanding the product development process using AI technology to other product development in the future,” it says. 

Kirin Holdings thinks that combined machine learning and simulation technology can be widely used in R&D for businesses. “Going forward, based on the knowledge gained from this research, Kirin Holdings aims to build a further foundation for research and development and deliver new value to its customers.”

Research background
FANCL is noted to have been calculating the removal rate of makeup stains since 2016.

The corporation took the initiative observing that as more women enter the workforce, there is a need for shorter work hours. This is why they are combining cleansing with facewash, thus, retroactively reducing environmental impacts.

On the other hand, Kirin Holdings has accumulated knowledge of machine-learning technology via its Brewing Takumi AI beer development.

Furthermore, Keio University’s molecular simulation technology was used to improve the prediction accuracy as makeup removal performance is “strongly influenced by the structure of the surfactant molecules used and the structure forming the surfactant and its adsorption state on dirt.”

“Keio University used a molecular simulation technique called the DPD method to simulate the structure and behavior of cleansing agent mixtures on a computer. After simulating more than 100 prescriptions, a correlation between the actual cleansing rate and the simulation results of how dirt is absorbed was found,” concludes Kirin Holdings.

By Venya Patel

To contact our editorial team please email us at editorial@cnsmedia.com