As its adoption grows, a diverse set of strategies is taking shape, with leading companies already commercializing materials designed with artificial intelligence and redefining what is possible in R&D.
Revenues for materials computing service providers are projected to reach a robust compound annual growth rate of 9.0% by 2035. The report also examines the transformative impact of the current rise in artificial intelligence and highlights numerous pioneering projects in the field of materials science. Furthermore, it demystifies the underlying technologies driving this digital transformation in R&D, offering readers a clear vision of the future of innovation in smart materials.
What is materials informatics?
In essence, materials computing leverages powerful data infrastructures and machine learning techniques to accelerate the design, discovery, and optimization of materials processing. By integrating data-driven methods throughout the R&D process—from hypothesis generation to data acquisition, analysis, and knowledge extraction—materials computing is transforming traditional workflows and enabling smarter, faster innovation.
Beyond predicting material properties, materials computing enables reverse design: starting with a set of desired properties and working backward to design the ideal material. This shift dramatically reduces the lengthy trial-and-error processes that have historically dominated materials development, making discovery faster, cheaper, and more specific than ever before.
However, materials computing presents unique challenges compared to other AI-driven sectors, such as autonomous vehicles or social media. Datasets are often sparse, high-dimensional, biased, and noisy, requiring specialized expertise to fully leverage their value. Successful approaches bridge the gap between materials scientists and data scientists, combining deep material knowledge with advanced analytics. When properly integrated, materials computing becomes an essential enabler, accelerating R&D and enhancing the impact of expert insights across the innovation process.
What will the situation of the sector be in 2025?
In recent years, the growing awareness of the need for digital transformation in R&D has accelerated the adoption of materials computing processes by players in the materials industry, from startups to established giants. The need for data-driven methods is becoming increasingly prevalent in the materials industry. Virtually all major players in the sector appear to have embraced materials computing in some way, whether through outsourcing, participating in consortia, or developing in-house programs.
With the rise of AI in 2023, interest in materials computing has only increased. Industry players told IDTechEx during interviews for their report that while adoption in the past tended to come from the bottom up within organizations, now the impetus is increasingly coming from executives eager to demonstrate the impact of AI on their businesses.
So far in 2024 and 2025, several impressive new players have emerged. Berlin-based startup Dunia Innovations, which focuses on materials discovery through physics-based machine learning and laboratory automation, launched last October with $11.5 million in funding. This March, Lila Sciences, a Cambridge, Massachusetts-based biotech venture capital startup, Flagship Pioneering, announced a $200 million seed investment to build its “scientific superintelligence platform and fully autonomous laboratories for life sciences, chemistry, and materials.” Notably, both Dunia and Lila have shown strong interest in heterogeneous catalysis for applications such as green hydrogen production, highlighting the potential of materials computing to influence sustainable development.
Materials computing activities at large technology companies have gained prominence since 2023. Microsoft's Azure Quantum Elements, which uses AI-driven selection and accelerated density functional theory simulations for materials development, has seen more published use cases across various materials fields with companies such as Johnson Matthews, AkzoNobel, and Unilever. Meta's fundamental AI research team also made a massive 110-million-point dataset on inorganic materials publicly available in 2024, hoping to foster materials discovery projects for applications such as sustainable fuels and AR devices. Over the next five years, the biggest challenges for established materials computing providers appear to be the growing interest in materials from large technology and AI companies, as well as the in-house development of materials computing platforms by materials companies themselves.
Author: Sam Dale, Senior Technology Analyst at IDTechEx
