NSF PCL-Test Bed: AI-enabled, Cloud-based Self-Driving Nanofabrication Node for Accelerating Semiconductor Research and Education

Award ID: 2607580

NSF PCL-Test Bed: AI-enabled, Cloud-based Self-Driving Nanofabrication Node for Accelerating Semiconductor Research and Education This project will address a longstanding challenge in semiconductor research and innovation, viz., the slow, costly, and difficult to reproduce processes involved in the development of miniscule, micro- and nanoscale electronic, photonic, and sensing devices due to the fact that many fabrication processes continue to rely on manual testing and expert judgment. The project aims to create a secure, remotely and broadly accessible laboratory that connects nanofabrication tools, measurement systems, structured data, and computing resources into a unified experimental environment. By using human-supervised AI-assisted workflows, the platform will accelerate learning from experimental data and dramatically reduce the time required to move from promising concepts to validated devices. This investment will strengthen U.S. leadership in semiconductor innovation, enhance national competitiveness, and support technologies essential to computing, communications, health, and national security. The project will also broaden access to advanced research infrastructure and provide hands-on education and workforce training for future semiconductor scientists, engineers, and technical professionals. The project will establish the Virginia Tech NSF Programmable Cloud Laboratory (PCL) Node for secure, remotely accessible, and human-supervised semiconductor nanofabrication research. This will be an AI-enabled, remotely accessible self-driving nanofabrication platform designed to accelerate semiconductor and nano device Design-Build-Test-Learn cycles. The platform will integrate nanofabrication instruments, characterization tools, structured data systems, and computational resources to accelerate iterative design, fabrication, measurement, and learning cycles. AI-assisted workflows will support experiment planning, process monitoring, measurement analysis, and iterative improvement of fabrication processes. The technical scope will include standardized experiment records, coordinated tool and metrology data, secure user access, workflow orchestration, and reusable process descriptions that improve reproducibility across users and projects. Representative validation workflows will include dielectric meta-optics, plasmonic biosensors, flexible bioelectronics, and thin-film lithium niobate quantum and telecommunications photonics. The intellectual contribution lies in advancing new knowledge on how to combine data driven models, physical process understanding, and human oversight to improve complex semiconductor nanofabrication workflows affected by variability, nonlinear behavior, and interactions across multiple length scales. The proposed methodology will develop an AI-driven Self-Driving Experimentation Framework that integrates physics-informed models, machine learning approaches, and human-in-the-loop oversight. The resulting infrastructure will establish the groundwork for faster and more reproducible semiconductor research, broader access to advanced nanofabrication capabilities, and workforce training at the intersection of fabrication, data science, and intelligent automation. This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.

NSF Program Director: Waleed Nasser