The University of Central Florida’s Institute for Simulation and Training (UCF IST) is leading the development of a semiconductor digital twin of the NeoCity Center for NeoVation in Kissimmee, Florida – aligned with BRIDG and other regional partners in the Osceola County Build Back Better Regional Coalition (BBBRC), sponsored by the U.S. Department of Commerce’s Economic Development Administration (EDA).
The Microelectronics Digital Twin (MeDT) framework provides a secure testbed for innovation and optimization, applying modeling and simulation techniques for semiconductor design, manufacturing, and workforce development. Initial use cases focus on the value proposition of modeling factory processes and operations for intelligence-based capacity planning for production optimization. UCF IST is leveraging its high-performance computing (HPC) capabilities to develop, train, and refine artificial intelligence / machine learning (AI/ML) algorithms for predictive maintenance – predicting tool failures based on historical performance and current/future fab conditions. Additionally, advanced capacity planning simulations seek to maximize throughput with efficient route planning – based on the current and future state of fab operations, resource availability, and business rules.
The MeDT application leverages advanced visualization to enable factory production, equipment, and process managers to view the current and historical states of the fab – including tool states, work in progress (WIP) data for all process flows, and key performance indicators (KPIs) such as moves/inventory (M/I), operational equipment effectiveness (OEE), alerts for predicted failures, etc. Additionally, users may run custom “what-if” simulations to experiment with different fab configuration scenarios and see the impacts on KPIs.
The MeDT application also serves as a dynamic and situated training environment for workforce development across the talent pipeline – from informal outreach with K-12 to formal training and certification programs at the community college, university, and industry levels. Example education and training scenarios include interactive simulations for tool operation, facility layout planning, factory processes, etc.
The MeDT framework is designed as a flexible and scalable architecture – adaptable across multiple facilities and across multiple applications, from electronic design automation (EDA) to manufacturing optimization, logistics / supply chain, and others. The architecture includes robust application programming interfaces (APIs) to pull and securely manage data from a variety of dynamic sources – i.e., IoT sensors, manufacturing execution systems (MES), fault detection systems, manufacturer specifications, scheduling software, and other enterprise systems – and will seek to align with emerging standards from SEMI and NIST for secure data sharing and interoperability.
