Conference 2014


Part 1: Introduction

Part 2: APC for Isolated Process Steps

  • Statistical process control (SPC)
  • Multivariate dependencies (sampling, correlation; multivariate SPC, principal component analysis PCA)
  • Limits of correlation and PCA (nonlinear dependencies, regression, virtual metrology, dynamic dependencies, dynamic partial models)
  • Control interventions (feed forward, feed backward, control scheme, metrics for control loop performance)
  • Application problems of control (control of delay time processes, smith predictor, virtual metrology)
  • Time variant processes (basic problem, predictive maintenance)

Klaus Kabitzsch

  • Affilation: Dresden University of Technology, Faculty of Computer Science
  • Klaus Kabitzsch holds a Diploma and a Ph.D. (1982) in electrical Engineering and communications technology from the Ilmenau Technical University.
  • He worked in the following years in the industry in the area of control of machines and equipment. After 1989 he worked at the universities in Leipzig and Dresden and focused his research on distributed systems, real time software and automation.
  • He became professor and head of the department of technical computer sciences in 1993 at the Dresden University of Technology. His current projects have their focus in the automation domain, component based software design and tools for data analysis and controller design.

Part 3: APC for Complex Processes

  • Mono-causal dependence - univariate analysis - SPC
  • Multi-causal faults - multivariate analysis
  • Linear or at least monotonic dependencies - statistical model sufficient
  • Nonlinearities and not normal distributed parameters - physical process model necessary
  • Examples: Plasma processes
    Main targets of plasma processing, recipe and process parameter,
    E-H mode at ICP plasma, process stability,
    Model based process analysis, DOE, plasma process properties and chamber matching,
    RF bias power and RF voltage and current - from generator to wafer,
    CP/CCP etch processes: transport issues,
    Electron density in coil zone provided by plasma model,
    Real bias power is influence by ICP power.

Michael Klick

  • Michael Klick is CEO of the Plasmetrex GmbH a plasma process control solution provider.
  • He got a diploma for technology of electronic devices in 1987 and a Ph.D. in plasma physics from Ernst-Moritz-Arndt-University Greifswald in 1992.
  • He has more the 15 yeas experience in nonlinear modeling of industrial RF Plasmas and development of plasma sensor systems for etch and deposition in semiconductor and PV manufacturing. He hold several patents in this area.
  • He has been working on the qualification of Process and Maintenance Personnel since 2003 and gives regular lessons at the for student about industrial plasma technology at the Ruhr-University Bochum.
  • Assessment, modeling, and RF and plasma chamber matching for plasma tools in manufacturing are the current key activities.

Part 4: APC on Fab Level

  • Why are fab-wide control solutions required?
  • Measurement aspects (dedicated measurement machines, in-situ sensors, upstream & downstream of process variables, yield, electrical parameters)
  • Fab level control engineering aspects (granularity, adjustments, definition of time, control hierarchies, control algorithms typically deployed at fab level: EWMA, MBPC, basic FDC algorithms to be found in fab-level implementations)
  • High-level requirements / attributes for fab-level APC solutions (availability, scalability, performance, flexibility, customizability, ease of use, security, integration into the environment)
  • Typical CIM system in the semiconductor industry (architecture / components)
  • CIM system integration and SEMI standards (PCS, E133, EDA, E120, E125, E132, E134, solutions for fab-level control, APC frameworks vs. applications)
  • Examples for fab-wide control solutions (CMP, lithography)

Roland Willmann

Roland Willmann holds a master in computer science from Vienna University of Technology. He is certified six sigma green belt and has successfully passed six sigma black belt education.

He worked in several software projects about computer aided testing, process documents and master data management, as well advanced process control in the semiconductor industry over twenty-five years. In 2000 he was the co-founder of CenterPoint GmbH and managed the development of a comprehensive software platform for advanced process control. Furthermore, he participated the development of the SEMI Interface-A standards.

In 2008 Roland Willmann became the product manager and lead software architect of the InFrame Synapse Process Control System (former CenterPoint's APC-Platform), where he also contributed with his APC-knowledge to several projects in the areas of photovoltaic, printed circuit boards, functional printing and semiconductory manufacturing.

Since 2014 he is the Engineering Manager at PEER Group GmbH in Dresden/Germany