November 20, 2019, Tokyo Japan – Kao Corporation and Preferred Networks, Inc. launch a collaborative project named “Kao X PFN Sebum RNA Project” with the objective of putting into practical use the technology which Kao has developed to monitor RNA (ribonucleic acid) in sebum.^*1

■ Initiative for practical applications of the sebum RNA monitoring technology

Initially, Kao and PFN aim to develop a beauty counseling service committed to skin conditions by applying AI technologies, such as machine learning and deep learning,^*2 to data obtained from RNA in sebum.
In this initiative, PFN will utilize the information obtained from Kao’s sebum RNA monitoring technology to develop a highly sophisticated prediction algorithm based on PFN’s machine learning and deep learning technologies. This will enable a better understanding of the internal conditions of the skin, and also assess future risks of skin damage. Further, this will pave the way for improvement and preventive measures of skin conditions by providing personalized beauty advice or skincare, based on genetic information. Some of these features are slated to begin in 2020 on a trial basis, with further improvements implemented based on user feedback.
In addition, Kao and PFN plan to conduct joint research into technology for improving early diagnosis of intractable diseases, including Parkinson’s disease, which has been increasing amid the aging society.

■ Structure of this project

【Kao】will collect approximately 13,000 types of RNA per person, measure RNA expression levels through Kao’s sebum RNA monitoring technology. In addition, Kao will obtain data concerning skin and other health conditions.
【PFN】will apply machine learning and deep learning technologies to train and build a prediction model using the sebum RNA expression and other data that can infer the skin conditions and biologic factors within the body.

*1   Sebum RNA monitoring technology is used to isolate RNA, which reflects day-to-day changes in conditions inside the body, from sebum, and analyze it. About 13,000 types of RNA’s expression levels can be obtained from a sebum sample in a non-invasive test (less stress on the body) using an oil blotting film.

*2   Deep learning technology allows machines to automatically extract features or patterns from a large amount of data. Deep learning has significantly increased the accuracy in certain tasks such as image recognition and voice recognition.

■Expectations for this collaborative project

Michitaka Sawada, President and CEO, Kao Corporation

“Kao is aiming to provide a practical method to accurately monitor biological information by applying our sebum RNA monitoring technology. To achieve this goal, we have high expectations that collaborating with PFN will greatly improve the accuracy and speed, thanks to PFN’s considerable experience with the utilization of AI technologies in the bio-healthcare field. Kao is focusing on social innovations that contribute to improved QOL under an ESG-driven management strategy. This project is part of that initiative, and we will contribute to the future society in collaboration with PFN.”

Toru Nishikawa, President and CEO, Preferred Networks, Inc.

“By combining PFN’s machine learning and deep learning technologies and accumulated know-how in RNA analysis with Kao’s research achievements in dermatology, productization expertise, and marketing capabilities, we expect to accelerate practical applications of products and services that utilize PFN’s technologies.”