Digital twin is a purpose-driven virtual representation of a physical object, system or a process that can provide intelligent feedback (eg; forecasting, optimization of parameters, root cause analysis, real time control) to the physical world with simulation, emulation, data analytics and AI modeling. With the advent of AI and Industry 4.0 this is becoming a critical component of Smart Manufacturing.
Machine Learning and Deep Learning AI models require Digital Twins which generally consist of three types of Digital Twins in semiconductor manufacturing: 1. Component Twins – digital representation of individual parts 2. Process Twins – digital representation of entire production facilities 3. Enterprise Twins – digital representation of entire manufacturing life cycles
There are key elements on the path to digitizing the IC manufacturing process to enable creation of process digital twins: 1) Sensors – Like FDC, IoT, Cameras, or cell controllers capturing unique behavior of the physical process; 2) Common platforms to provide access and link data from the entire manufacturing value chain; 3) Analytics – Advanced analysis and modeling tools able to be leveraged in manufacturing for things like Virtual Metrology; 4) Knowledge Catalogues & Data models (digital twins) that allow meaningful and relevant insights to be drawn from the digital data; 5) Direct CPS/CPPS connections: Can’t ACT if you don’t change tool state or decision process; and lastly; 6) Site-to-site enterprise connectivity: Must integrate data across the entire supply chain.
To achieve the goal of AI powered Smart Manufacturing Provide a scalable digital twin data platform, common and continuous connectivity to realtime physical processes, and the ability to leverage current and emerging SEMI standards as well as streaming and custom containerized data for direct AI/ML applications. This must be an Automated system that can effectively connect manufacturing and supply chain data together in a digital twin with minimal human involvement to enable faster decisions and better control systems. Ability to automatically create and deploy AI and Machine Learning models with high accuracy and deploy to an edge control platform to enable leveraging factory data for Al/ML applications. This requires a scalable architecture that enables high performance processing to execute all of the above at the required levels of automation while maintaining Root of Trust and Traceability for all materials and components in the supply chain.
