8/16/2023

SKINLY study: Skin research meets artificial intelligence

It is hard to imagine our everyday lives without artificial intelligence (AI): whether smart homes or face recognition, navigation apps or spam filters, voice assistants or online shopping recommendations – the algorithms used for this purpose are pretty much found in almost all areas of life. Beiersdorf also uses algorithms specifically to learn more about skin: With SKINLY, one of the world’s largest skin studies with consumers, the company is using the possibilities of the Internet of Things (IoT) and artificial intelligence – and thus opening up a whole new dimension of skin science. In this interview, Sven Clemann, Lab Manager for “Advanced AI and IoT” and Head of the SKINLY study, and Sören Jaspers, Head of Digital Pioneering at Beiersdorf Research & Development, provide insights into this special study and explain why artificial intelligence alone is not enough.

Sven, SKINLY sounds kind of “cute” – but you’ve really created something big and unique. Can you tell us a bit more about what SKINLY is all about?

So much more than a measuring device: Physicist Sven Clemann is building up an enormous wealth of knowledge with SKINLY.

Sven: SKINLY is one of the largest skin care studies in the world. What makes it special is that we actively involve consumers via digital tools. “Consumer Connectivity” and “Data Thinking” are the key words here. With this study, we are continuously expanding knowledge about the skin and creating data validity for scientific evaluation. For this purpose, we have specially developed an IoT measuring device and a supplementary app that all study participants can use via their smartphones.

On the one hand, we are thus advancing digital skin research, and on the other hand, we are enabling do-it-yourself data collection from home via SKINLY, which creates real added value thanks to artificial intelligence. A real novelty in the field of skin and skin care research!


How exactly do the skin measurements work?

Sven: The measurements are taken directly at our consumers’ homes via the measuring device mentioned. The SKINLY measuring device has a moisture sensor and a special camera with three different light sources: direct light, fluorescence stimulating light and polarized light. Consumers use it twice a day to measure individual skin characteristics on the cheek, eye wrinkle, forehead and forearm – i.e. skin age, wrinkle depth, skin tone, tints of the skin and impurities. The results are analyzed and recorded in the associated app, which also displays the participant’s individual skin report after the measurement has been taken.

In addition to the daily measurements, our study participants can also use the app to record data on topics such as sleep, nutrition, skin peculiarities or the menstrual cycle.

SKINLY is not just providing snippets, but accompanies consumers continuously.

Furthermore, the system also takes into account the respective climatic conditions of the people taking the measurements. That itself may not sound that spectacular. What is really impressive, however, is the unique ecosystem that we build up in the course of continuous measurements. This is real “machine learning” based on artificial intelligence.

What makes it so special?

“Better data quality enables the development of better products,” says Sören Jaspers.

Sören: The ecosystem we are building in this study is continuously learning with each measurement. The more data we have, the better we can understand the external influences on the skin. It’s not just about the sheer quantity, it’s primarily about the quality of the data. Only with truly valid data can we ensure maximum informative value and draw meaningful conclusions for our research and development work. Initial evaluations of our data have shown us that many consumers are not very good at assessing their own skin and therefore may not use the optimal care products for their individual needs.

SKINLY creates a clear advantage here: we receive real, reliable data – directly and under very real conditions. With the help of artificial intelligence, we can develop and recommend better products for consumers’ needs – and thus deliver tangible added value for our consumers in the long term.

Artificial intelligence, machine learning – it all sounds very futuristic. So are systems gradually replacing humans in this work?

Sören: Fortunately, that’s not true! We are not handing over the work, it is rather a valuable, additional and very effective tool we can use. Artificial intelligence is not a black box that does everything on its own. Because data science based on artificial intelligence always needs humans as well, who feed in high-quality data, develop models and then analyze, process or optimize them. Good artificial intelligence always learns from people. In other words, it’s about the perfect symbiosis of humans and the machine, because the application of AI is only as smart as the developer behind it. For Beiersdorf, this means that our experience is immensely important and gives us the decisive advantage in skin research.

Only the combination of skin science, product experience, and digitization expertise creates the decisive added value here. It helps us to read the right things from the data. And in this way, AI can and will increasingly be an important source of innovation and contribute to increasing efficiency, for example also by noticeably supporting formula optimization.

Back to SKINLY. What data volume are we talking about today? How many measurements have already been done and where?

Sven: The SKINLY study has been running since 2020 and until now more than 7 million skin measurements have been taken worldwide. More than 17,000 consumers aged between 18 and 80+ years have participated in SKINLY so far. This has generated more than 800 million data points and around 90 million high-quality skin images. This is really an incredible amount that can only be processed using digital tools. What is particularly exciting for us are the different locations. We started in Germany, but we have also conducted organized studies in Thailand, South Korea and China, for example. Thanks to the measurement data, we can see exactly what influence a humid or dry climate has – or a rural versus an urban environment, for example in megacities like Seoul or Shanghai. It was also important for us to take a cross-product as well as cross-brand approach so that we could obtain the most comprehensive results possible. Since many participants also take the measuring device with them when they travel, we now have measurement data from more than 50 countries.

Beiersdorf scientists specially developed the state-of-the-art IoT tech device and the associated app for the skin study.

Are there already new products that have been created based on the collected data? And how does the overall future of the study look like?

Sören: To be honest, it’s still too early for that. But we have already gained very exciting insights into the performance of our products with consumers, for example about the impressive effect of W630 and the associated consumer reactions. In the future, we expect to see, among other things, more modern skin typing that is easier for consumers, as well as new insights into the optimal combination of complementary products. Artificial intelligence is also increasingly being used in product development itself. Sustainable ingredients, for example, are becoming ever more important, and AI is helping us to develop formulas faster and replace existing ingredients with sustainable alternatives.

The pace of development of AI and digitalization in general is enormous – and there will certainly be a lot more happening here in the area of research and development, which will open up new opportunities for us scientists and ultimately also for consumers. Future innovations will arise in particular at the interface and the optimal fusion of digitalization, skin biology and product development.

Sven: Talking about SKINLY, we will of course continue to generate high-quality data – and do so worldwide. Another organized study in the next focus country is already planned. We see a great opportunity and ultimately competitive advantage in the field of data science. Nevertheless, the core is and will remain our excellent skin research expertise, as this will allow us to effectively use AI for us and shape the skin care of tomorrow – not only for, but also with our consumers.

Sven and Sören, thank you for these exciting insights.

Related Links

About the editor: Kathrin Erbar

Kathrin Erbar

Kathrin takes us on a journey to the fascinating field of research and development at Beiersdorf. Before exploring Beiersdorf’s DNA, she was doing the communication for HR related topics, such as diversity, leadership or New Work. She also used to be responsible for financial communications at Beiersdorf for several years.

3 / 8 Open Positions
Ready for the challenge? Then apply now!
Print
Here you can print the page.