AI and machine learning methods can be used across a battery entire life, from material discovery and cell testing to second-life assessment, to enhance new developments and diagnostic capabilities. IDTechEx report, “AI-Driven Battery Technology 2025-2035: Technology, Innovation and Opportunities“, explores predictions and innovation possibilities within the sector over the next decade, stating that employing AI within the battery sector could bring about faster cell development time, improved yield and quality control, and the future development of new cell chemistries and battery formulations.
Development, manufacturing, in-life, and end-of-life
AI and machine learning could have a role to play in every stage of a battery’s life. The initial development stage could see material discovery, high-throughput cell screening, and temperature and pressure simulation all being assisted by AI technology. The manufacturing processes for AI applications that follow could then include defect identification, parameter optimization through regression, and quality control.
During the battery’s active in-life stage, AI and machine learning can be used for maintenance of the battery, to prolong its life and give it the best chance at reaching optimal performance. Charge-discharge protocol optimization, hazard identification, state-of-health (SoH) calculation, charge calculations, and defect detection will all be able to ensure safety is prioritized, such as when a battery is active within an electric passenger vehicle.
SoH calculations, defect identification, and second-life assessments are the processes then required to be carried out at the end of a battery’s life cycle. AI could provide the most accurate information to determine which applications it could be suitable for in a second-life run.
Energy densities, safety, and costs of rechargeable batteries
The main issues surrounding rechargeable batteries include challenges with energy density, the safety of liquid electrolytes, and the extraction of critical materials. Liquid electrolytes have a tendency to become dangerous under high temperatures and pressures, bringing about an opportunity for AI and machine learning to perform in-life diagnostics and monitoring, to help maintain a level of safety and assurance that, should any problems begin, they are detected early.
The limited supply of critical materials such as lithium, nickel, cobalt, and copper can pose difficulties for battery manufacturers, along with the high extraction costs and negative environmental impacts. These factors associated with sourcing highly sought-after materials incentivize battery repurposing and recycling, which can also be assisted through machine learning techniques.
AI in electric vehicle batteries
Electric vehicles are predicted to be one of the largest markets for AI in batteries. The possibility for AI to benefit and enhance the many stages of a battery’s life will, in turn, allow for better development, testing, and management of EVs. New battery chemistries and structures will be necessary for improving energy density, charging, and safety, all of which EV manufacturers aim to prioritize. The ability of AI to explore a range of chemistries efficiently and effectively could mean improvements are seen much faster and with less need for trial and error.
The assurance that the intelligence of AI and machine learning can provide could also be unmatched when it comes to achieving strong diagnostics, as levels of safety can be achieved that weren’t previously possible. Battery management systems will also enable users to access data as it is being analyzed and interpreted, for greater involvement in what is happening with the vehicle.
IDTechEx’s “AI-Driven Battery Technology 2025-2035: Technology, Innovation and Opportunities” report provides details on different machine learning approaches, material discovery, cell testing and assembly, battery management system analytics, and second life assessments, along with forecasts for the uptake of new technologies.
