Equipped with advanced clinical testing equipment and collaborative workspaces modeled after Lubrizol’s Barcelona clinical testing facility, the Shanghai Institute offers a variety of AI-empowered R&D capabilities.
While Lubrizol has a long history of success in developing new ingredients using classical scientific disciplines, the company recognized a few years ago that these approaches were becoming increasingly limited in their ability to handle the growing volume and complexity of data generated by its research. “We arrived at the use of new tools out of necessity,” said Barcelona-based Sr. Global Technology Director Raquel Delgado.
While Lubrizol has developed new active ingredients for the beauty industry using classical science disciplines like biology, molecular biology, and chemistry for the past roughly 25 years, a few years ago, the company began to increasingly see several realities emerge. One, traditional methods alone were no longer sufficient to handle the sheer volume and complexity of data generated by its research. Second, data science and analytics tools and experts were “blooming,” Delgado said. “New data analysts were showing up with the capability to process huge amounts of data in a different manner, essentially generating web-based applications, so generating algorithms that help us to process huge amounts of data that we were collecting.
Reimagining a blockbuster
In addition, Lubrizol saw that AI tools held the potential to significantly enhance its product development, particularly in its efforts to create a more potent version of its blockbuster anti-wrinkle peptide, Argireline, which debuted in 2001. The company sought to “generate, years later, a much more powerful peptide,” as Delgado explained.
The research would expand the application of this popular ingredient, which has long been known as a topical alternative to Botox injections. The project eventually led to the development of the enhanced formulations Argireline Amplified and Argireline YOUth peptides. The latter is designed to be oil-soluble thanks to the company’s patented LipoClear inverse micelle delivery system. These new formulations target both mature and younger demographics, with Argireline YOUth specifically addressing the concerns of Gen Z and millennial women.
Computer vision unlocks deeper patterns
To fully assess the efficacy of these new formulations on a diverse range of skin types, Lubrizol turned to a technology that continues to undergo something of a renaissance: computer vision. “We use this technology to translate into parameters the emotions that our products created in the skin of the volunteers,” Delgado said. This led to the development of SmiLearning, a proprietary tool that analyzes facial expressions to quantify consumer satisfaction. Volunteers using the active ingredient treatment showed a 22% bigger smile compared to the placebo group. The research behind SmiLearning involved recording and analyzing thousands of frames from videos of volunteers testing Lubrizol’s products at their Beauty Research Institutes in Barcelona and Shanghai.
Quantifying the lifting effect
Lubrizol also tapped AI-driven insights to inform the design and development of Uplevity e-Lift peptide. This peptide, inspired by the effects of microcurrent devices, aims to deliver a similar lifting and firming effect through a topical application. “We assessed how the skin was progressively getting a lifting aspect, but we needed to quantitate it. We needed to provide numbers. We needed to provide evidence,” explains Delgado.
Traditional research equipment produced overwhelming visual data. Delgado describes the methods as “just providing different arrows that were pointing up or down.” She continued: “Imagine one face crowded with arrows of different lengths.” Traditional approaches also tend to result in an unmanageable amount of data across multiple checkpoints and numerous study participants. To address this challenge, Lubrizol again turned to computer vision. The computer vision team transformed the complex visual data “into mean volumes and angles,” allowing Lubrizol to quantify the lifting effect. This AI-driven approach enabled Lubrizol to “assess in each volunteer, before and after the treatment.” The technique allowed the researchers to see “how [participants’] skin was moving.” The technique was able to “provide the objective benefit with scientific parameters. “Long story short, without such tools, this was never going to happen,” she said.
Fostering an entrepreneurial approach to AI
AI initiatives may be hot, but many AI projects fail to deliver on hopes. A 2023 Harvard Business Review article estimated that eight of ten such initiatives fail while also noting that there are strategies organizations can take to improve the odds. That includes focusing on priorities such as business needs, culture and talent. In Lubrizol’s case, the company has a dedicated corporate innovation team focused on exploring and implementing AI technologies across its business units, including automotive, industrial, and human health sectors.
Lubrizol’s corporate innovation team supports business units in developing and implementing AI initiatives. This team structure allows business units to prioritize innovation. Lubrizol’s AI adoption strategy prioritizes speed and adaptability. The innovation team works with business units to ensure AI initiatives meet market demands while adhering to a company-wide AI strategy.
In addition, Lubrizol invests in upskilling existing employees in AI and deep learning through training programs and opportunities to work on AI-driven projects. The company attracts new talent by offering opportunities to work with emerging technologies.
Laying the groundwork for the future of skincare
Being an early AI adopter can have clear advantages when it comes to competition. “I would say that we wanted to be ahead of the curve of technology,” Delgado said. Lubrizol, active in a range of industries beyond beauty — including transportation, industrial, consumer, advanced materials and life sciences — recognizes the potential for cross-divisional benefits from AI initiatives. “Bear in mind that [AI] capabilities do not stop with beauty.”
Delgado highlights the established role of in vitro cell-based assays for assessing beauty product performance, but also underscores the need for complementary technologies that can capture the nuances of how products perform on diverse skin types. “We need to understand how it performs in the different skin types,” she explains, emphasizing the strategic importance of having advanced research facilities like those in Barcelona and Shanghai. These facilities, equipped with AI-powered tools, can help bridge the gap between in vitro testing and real-world application, offering new ways to visualize and quantify product performance across diverse populations.
Building a collaborative future
Lubrizol understands that while equipment exists to evaluate beauty product aspects like moisture, wrinkles, and lifting, these tools have limitations. The company’s vision is to tap the expertise of data scientists and computer vision experts to unlock new levels of understanding in beauty research. This endeavor requires integrating team members with diverse expertise and fostering a culture of collaboration. “There are commonalities that help to integrate the team, where everyone has their own specific expertise,” Delgado explained.
The company is also dedicated to internal talent development through training programs and opportunities to work on AI-driven projects. It also actively recruits young talent. “This new generation is eager to learn what we have been doing so far,” Delgado said, “But the team that we have in the company as well is eager to connect with the new opportunities.” Ultimately, Lubrizol believes that by fostering continuous learning and embracing new technologies, it can drive meaningful innovation in the beauty industry and beyond. As Delgado concluded, “I really believe that…this is how we are going to be able to solve a lot of problems in the future, combining this data with the classical science.”
Tell Us What You Think!