The primary objective of this lecture is to share the latest technology from the University of Michigan in the area of Fab-wide virtual metrology, feedback control and performance of industrial Ethernet applied to networked control, diagnostics and safety. In this presentation, an approach to factory-wide deployment of Virtual Metrology (VM) will be presented. The speaker will detail the formulation of the VM problem for a multi-step multi-input (process variables) multi-output (quality or metrology variables) process environment which represent a typical semiconductor manufacturing 'fab-wide' environment. The methodology for deployment of the VM model will be presented along with results of applying the methodology to simulated process data. Dr Moyne will explore the results achieved in Reconfigurable Manufacturing Systems (ERC-RMS) and will identify the key factors that should be considered in evaluating industrial Ethernet performance.
(1) Fab-wide Virtual Metrology and Feedback Control
In today’s semiconductor industry most processes apply simple SPC techniques or run-to-run (R2R) adaptive control utilizing the sampled metrology (measurement) data for lot-to-lot control. However as Moore’s law drives us towards increases in wafer size and decreases in device size, wafer-to-wafer (W2W) control will be the only choice to cope with the demanding situation. W2W control can only be achieved when metrology data is available for every wafer. Unfortunately issues such as impact on throughput, increase in cycle time, delay in measurement for feedback and cost of integrated metrology make achieving this capability impractical in many process environment. Recently virtual metrology (VM) techniques have been suggested as an alternative to 100% wafer measurement to support wafer-to-wafer W2W control. VM is realized by utilizing the process data from the underlying tools; this data is generally collected in real time for fault detection purposes. A typical fault detection and classification (FDC) system collects tool data (often referred to as process variables) on a R2R basis and this enormous amount of data (involving hundreds of variables) can be used for VM purposes. The VM systems utilize the data and provide online prediction of the metrology variables, thus enabling W2W process control. Researchers have analyzed a few processes for VM application but mainly these efforts are stand-alone in their nature and are seldom integrated with the process control application. In this presentation an approach to factory-wide deployment of VM will be presented. We will detail the formulation of the VM problem for a multi-step multi-input (process variables) multi-output (quality or metrology variables) process environments which represent a typical semiconductor manufacturing “fab-wide” environment. The methodology for development of the VM model will be presented along with results of applying the methodology to simulated process data.
(2) Performance of Industrial Ethernet Applied to Networked Control, Diagnostics and Safety
As industrial Ethernet becomes more prevalent on the manufacturing floor it is increasingly being considered as a universal networking solution. While the capability of Ethernet has been proven acceptable in high level manufacturing environments, issues of Ethernet performance must be considered as we push this technology into domains of control, diagnostics and safety, and into applications such as I/O, motion control, fail-to-safe, and security. The University of Michigan’s Engineering Research Center for Reconfigurable Manufacturing Systems (ERC-RMS) has been exploring the applicability of Ethernet to the factory floor. Results include: - An identification of node software performance as a dominant factor in evaluating overall Ethernet system performance. - An quantification of the delay overhead associated with common factory Ethernet protocols such as UDP, VPN and OPC for data transport, security and diagnostics respectively. - A comparative evaluation of common industrial Ethernet protocols of EtherNet/IP and PROFINET. - Development of a cost-based evaluation methodology for determining the need for partitioning of networks across functionalities of control, diagnostics and (especially) safety. This presentation explores the results achieved in the investigation of industrial Ethernet at the ERC-RMS and identifies key factors that should be considered in evaluating industrial Ethernet performance.
About the Speakers
Dr James Moyne is an Assistant Research Scientist with the University of Michigan. He is Director of Integrated Manufacturing Process Automation and Control Technologies (IMPACT) group. His research interests include computer integrated manufacturing, specifically in the VLSI and flat panel display manufacturing arenas, process and equipment control (supervisory, inter-process, run-to-run, and real-time/in-situ control design), cell automation, data modelling, database design and development, communication and networking protocol design, sensor bus system modelling and standardisation, and factory architecture and design.
Senior management, R&D managers, engineers, system integrators from the logistics, electronics and precision engineering industries; researchers, academic staff and students will find this lecture very informative and beneficial.
Registration Admission is free. Seats are available on a first-come, first-served basis. All are welcome.
Pre-registration for the lecture is necessary.
To reserve a seat, please include name, designation, email address, company name in your registration request via: eventsPostmaster@SIMTech.a-star.edu.sg
For technical enquiries: Mr Goh Kiah Mok, Email: kmgoh@SIMTech.a-star.edu.sg