The Paul Drude Institute for Solid State Electronics (PDI) in Berlin, Germany is collaborating with scientific technology company Bizmuth MBE Ltd. to integrate intelligent automation into the growth of semiconductor materials. The six-month partnership, running from June to December, is among the first initiatives in Europe to apply large language models (LLMs) and multimodal AI to the autonomous control of molecular beam epitaxy.

PDI has more than 30 years of expertise in molecular beam epitaxy and semiconductor materials, and the institute operates 11 molecular beam epitaxy systems at its Berlin facility.

The collaboration will focus on developing AI-driven control software for MBE systems, designed to automate adjustments during the growth process in real time. Unlike existing AI-assisted tools, which provide feedback but leave decision-making to operators, this system will operate autonomously. The software will run entirely on local edge hardware, ensuring data security and independence from cloud infrastructure. The project aims to improve the reproducibility, efficiency, and scalability of semiconductor research, while reducing material waste and system downtime.

The growth of high-quality gallium nitride (GaN) has already been successfully established within the semiconductor industry and will serve as a reference system for the project. The AI software will be implemented and tested on GaN at PDI before being applied to more complex materials systems, where conventional trial-and-error approaches for determining growth parameters are often too time-consuming and costly.