Artificial Intelligence with Open and Scaled Data Sharing in Semiconductor Manufacturing – Workshop Report

Published

November 18, 2025

Author(s)

Sthitie Bom, Jim Davis, Bruce Kramer, Gregory Vogl, Said Jahanmir, Don Ufford

Abstract

The Workshop sponsored by the National Science Foundation (NSF) (NSF award 2334590, "Artificial Intelligence with Open and Scaled Data Sharing in the Semiconductor Industry") and supported by the National Institute of Standards and Technology (NIST) was organized to evaluate the promise of scaled Data Sharing in data and operational ecosystems. The Workshop conducted an unprecedented industry demonstration of how a coalition of manufacturers in the semiconductor industry merged their data to seek new operational opportunities, improve the predictive capabilities of AI/Machine Learning (ML)/Digital Twin (DT) systems, and return benefits to individual manufacturers. The Workshop was conducted as a series of five Roundtables between July 2023 and July 2025, centered on an agreed upon AI/ML virtual metrology use case for which the data and modeling execution requirements could be considered in detail with ecosystem results benchmarked. The industry coalition committed to testing ecosystem Data Sharing as a business priority. This emphasis highlighted the importance of a Data-First Strategy. A Seagate/UCLA project team benchmarked technical value points by executing data processing workflows to refine raw data from multiple machine site sources into qualified AI-Ready Data and then build and test collaborative AI/ML metrology models. Benchmarking was done by comparing performance with cross machine data vs. the baseline of individual machine performances. The result is a scalable data process that maximizes the value of Data Sharing by maximizing many value points with data processing, aggregation, usage, and model performance when acting together in an ecosystem. The Workshop demonstrated that manufacturers, even competitors, can agree on a meaningful process for Data Sharing to achieve significant factory and company operational benefits. Factories and companies can act together to prepare data and build collaborative AI/ML models, e.g., for wafer processing, while acting separately on their own products and applications. Importantly, training data from similar operations at different factories led to more robust modeling, opened cross operational opportunities and insights, and suggested new data service possibilities, as well as a Grand Vision for AI/ML systems that draw value from cross company data. All of this was accomplished cost effectively by working with the industrial participants as a Collaborative Data Ecosystem (CDE) using existing manpower and known technologies.

Citation

Advanced Manufacturing Series (NIST AMS) - 100-72

Report Number

100-72

NIST Pub Series

Advanced Manufacturing Series (NIST AMS)

Pub Type

NIST Pubs

Download Paper

https://doi.org/10.6028/NIST.AMS.100-72

Local Download

Keywords

Advanced Manufacturing, AI, Artificial Intelligence, Collaborative Data Ecosystem, Data Sharing, Machine Learning, Manufacturing, ML, Semiconductor Manufacturing, Smart Manufacturing

Citation

Bom, S. , Davis, J. , Kramer, B. , Vogl, G. , Jahanmir, S. and Ufford, D. (2025), Artificial Intelligence with Open and Scaled Data Sharing in Semiconductor Manufacturing – Workshop Report, Advanced Manufacturing Series (NIST AMS), National Institute of Standards and Technology, Gaithersburg, MD, [online], https://doi.org/10.6028/NIST.AMS.100-72, https://tsapps.nist.gov/publication/get_pdf.cfm?pub_id=960553 (Accessed November 20, 2025)