Artificial intelligence (AI) has the potential to cut in half the time required for developing new battery materials, according to executives in the battery industry. Mathias Miedreich, CEO of Umicore in Belgium, partnering with Microsoft, believes AI will have a profound impact on the discovery and commercialization of innovative battery materials.
In an interview with the Financial Times, Miedreich outlined the ambition to significantly reduce the time to market for new products. He stated, “The ambition is, when you have a certain product that you want to develop, to cut the time to market by half…we are doing now the first concrete projects to see if this is feasible.”
Umicore, a supplier of materials for car batteries used by major automakers like Volkswagen, BMW, and Stellantis, has been collaborating with Microsoft for the past two years. Miedreich expects AI to shorten the research stage for developing new battery materials from five or six years to just two years.
The partnership has enabled the filing of AI-enabled patents, with the technology showing particular promise in breakthroughs for solid-state materials and disorder rock salts. These advancements could significantly improve the driving range of electric vehicles and open up new applications for batteries.
Battery producers are optimistic about the transformative impact of AI on material discovery, drawing parallels with the pharmaceutical industry’s use of technology to expedite drug development.
Siddharth Khullar, a former Apple AI researcher leading machine learning at Northvolt, the Swedish battery manufacturer, emphasized the substantial speed-up AI could bring to different phases of the process. He likened the battery development process to a recipe with multiple ingredients, stating, “With AI, you can start to expand the comfort zone of where you look.”
Microsoft recently reported successful results from its collaboration with the Pacific Northwest National Laboratory, demonstrating that its algorithm could narrow down 32 million materials to 18 promising candidates for battery use in just 80 hours.
Despite the positive outlook, Miedreich cautioned that the availability of computing power might pose a challenge to achieving all breakthroughs. Additionally, testing new materials in real-world conditions and ensuring the safety of batteries produced with these substances remain critical challenges.
Nonetheless, with the United States perceived to lead in AI over China, experts believe it presents an opportunity for the West to catch up with Asian rivals dominating the global battery industry. James Frith, principal at Volta Energy Technologies, expressed the potential for the U.S. to use AI to optimize battery materials, reducing reliance on China.