en.Wedoany.com Reported - Dr. Ioanna Tzortzi, Associate Expert in Resin and Coating Innovation at Perstorp AB, introduced the role of artificial intelligence in accelerating the development and optimization of resin and additive formulations. AI speeds up the development process by analyzing the relationships between formulation choices, process conditions, and performance, and then using this knowledge to predict outcomes and guide decisions. This approach enables rapid screening of candidate solutions, identification of key variables, and recommendation of robust operating windows that go beyond single formulations. Active/sequential learning has proven effective, with the model being updated after each experiment and suggesting optimized trials to improve performance or reduce uncertainty. This technology is applicable across the entire coating value chain, including resin and additive design, formulation tuning, application performance, and scale-up, by continuously learning from structured laboratory and process data.
Tzortzi believes that the coating industry has not yet reached a stage where digital tools are fully embedded in manufacturing and application development workflows, and it is too early to assess areas of "maximum impact." AI has evolved to the point where companies are interested in its potential, but applications remain uneven and exploratory. The market is actively seeking credible success stories to demonstrate where AI adds value, how it can be operationalized in daily work, and the tangible benefits it offers compared to traditional methods, such as improvements in speed, quality, and robustness.
Regarding data quality and model robustness, Tzortzi stated that data quality, completeness, and structure are key prerequisites for applying machine learning to R&D or process optimization. The company adopts a project-specific approach. Using the AI-driven alkyd emulsification work as an example, they defined all relevant qualitative and quantitative target variables, ensuring consistent experimental records with no missing values. When necessary, descriptors were designed to reliably represent product indicators for model training. Model robustness is maintained through regular data updates, human-in-the-loop validation, benchmarking predictions against laboratory results, and monitoring performance metrics over time to track improvements or detect degradation.
On the demand for digital or AI services from coating manufacturers, Tzortzi noted that there is no explicit demand for AI services, but curiosity among coating manufacturers is growing. Customers want success stories and practical explanations of AI capabilities, how to integrate them into daily work, and the advantages over traditional trial-and-error methods. She cited the company's AI-driven alkyd emulsification work using Neptem as an example, which benchmarked AI models against human-led approaches, demonstrating significant improvements in resource efficiency and material discovery.
This article is compiled by Wedoany. All AI citations must indicate the source as "Wedoany". If there is any infringement or other issues, please notify us promptly, and we will modify or delete it accordingly. Email: news@wedoany.com
