The materials science industry is highly competitive. There are large changes in various sectors, such as an electric and autonomous future for vehicles, geopolitical & legislative developments, and new players offering innovative materials looking to gain market share. The key to future success will rely on commercializing R&D activity and remaining agile to market needs, two reasons that developing a materials informatics strategy is imperative.

Materials informatics (MI) is based on using data infrastructures and leveraging machine learning solutions for the design of new materials, the discovery of materials for a given application, and optimization of how they are processed. MI can accelerate the “forward” direction of innovation (properties are realized for an input material) but the idealized solution is to enable the “inverse” direction (materials are designed given desired properties).

Crucially this allows the time for R&D to market to be reduced for property optimization, new material discovery, or in response to supply chain shocks.

IDTechEx has been reporting on the topic of materials informatics for multiple years. Recent articles have discussed what has changed in recent years and the considerations to achieve a digital transformation. For more information see the leading market report on the topic, “Materials Informatics 2022-2032”.

Firstly, it is important to note that this field is still in its infancy with many of the key advancements still at an academic level. The technology and applications are still rapidly evolving and, although success stories are expanding, the real value will not be seen for many years. Most of the exciting progressions are in developing AI appropriate for materials science data and integrating domain knowledge, but the reality for most companies is systematically getting data in usable formats. Improvements in automated laboratory equipment and the emergence of quantum computers will only augment this revolution.

Some of the more bullish disruptive players believe there will be some major casualties that will cause laggards to wake up too late and that fast-followers will pay a premium to catch those early adopters. IDTechEx do not view anything as dramatic happening but embracing a data-centric R&D future will increasingly pay dividends in the decades to come.

It is often questioned why this trend has not happened sooner; equivalent approaches in the pharmaceutical industry are continuing to improve but are largely well-practiced. The are several reasons including materials science being a less homogeneous field with less value held in IP, as opposed to know-how, and a reasonable amount of skepticism and hesitation to overcome.

The number of companies providing external services is rapidly expanding and many notable organizations are engaging with these companies or consortia as well as building their own in-house capabilities. IDTechEx view the most promising applications are with thin film materials and liquid formulations, the latter is certainly where most of the commercial activity is seen for polymers, coatings, lubricants, and electrolytes. That is not to say we will not see increasing results and adoption elsewhere, there are some early wins in metal alloys, heterogeneous catalysts, superconductors, and many more, see image. Rather than considering the material families, it can also be beneficial to look at problems that this has seen success in such as screening for a band gap, mapping a phase diagram, or reducing your computational load. The market report compares the different strategies, interview-based player profiles, and numerous case studies.

The risk will be disillusionment and frustration with the results. Many of the case studies to date are cherry-picked and as a result not appropriate for all the real-world variables; It is, therefore, important for materials and chemical companies to pick relevant projects and demonstrate small wins internally.

IDTechEx view the end goal for any materials or chemical companies to have the capabilities in-house which will require a diverse multi-disciplinary team rather than a data scientist silo. The best approach is to work on a real in-house project that is generating a large amount of collectible digital data (this could be optimization or early-stage R&D) and is a multi-dimensional problem. The company could use an external provider to support this, e.g., with training, research projects, or using their platform, but for the long-term building in-house tools and know-how is essential. This is a long journey, but one that every chemical and materials company needs to take.