Industrial Automation

Effective RFID-based object tracking for manufacturing

by James Brusey

Weighted Windowed PLS Models for Virtual Metrology of An Industrial Plasma Etch Process

by Shane Lynn

International Conference on Industrial Technology, Vina del Mar, Chile, Mar 2010.

Virtual metrology is the prediction of metrology variables using easily accessible process variables and mathematical... more Virtual metrology is the prediction of metrology variables using easily accessible process variables and mathematical models. Because metrology variables in semiconductor manufacture can be expensive and time consuming to measure, virtual metrology is beneficial as it reduces cost and throughput time. This work proposes a virtual metrology scheme that uses sliding-window models to virtually measure etch rates in an industrial plasma etch process. The windowed models use partial least squares (PLS) regression and a sample weighting scheme to combat the effects of both process drifts due to machine conditioning and process shifts due to maintenance events. An industrial data set is examined and the weighted windowed PLS models outperform global models and non-weighted windowed models

E-learningHUB of production automation

by Jari Jussila

E-Learning Environment of Building Automation

by Jari Jussila

The Cambridge Packing Cell: A Holonic Enterprise Demonstrator

by James Brusey

Safe Programming Languages for ABB Automation System 800xA

by Markus Borg

The old master thesis should be here to make me feel... uncomfortable.

More than 90 % of all computers are embedded in different types of systems, for example mobile phones and industrial... more More than 90 % of all computers are embedded in different types of systems, for example mobile phones and industrial robots. Some of these systems are real-time systems; they have to produce their output within certain time constraints. They can also be safety critical; if something goes wrong, there is a risk that a great deal of damage is caused. Industrial Extended Automation System 800xA, developed by ABB, is a realtime control system intended for industrial use within a wide variety of applications where a certain focus on safety is required, for example power plants and oil platforms. The software is currently written in C and C++, languages that are not optimal from a safety point of view. In this master’s thesis, it is investigated whether there are any plausible alternatives to using C/C++ for safety critical real-time systems. A number of requirements that programming languages used in this area have to fulfill are stated and it is evaluated if some candidate languages fulfill these requirements. The candidate languages, Java and Ada, are compared to C and C++. It is determined that the Java-to-C compiler LJRT (Lund Java-based Real Time) is a suitable alternative. The practical part of this thesis is concerned with the introduction of Java in 800xA. A module of the system is ported to Java and executed together with the original C/C++ solution. The functionality of the system is tested using a formal test suite and the performance and memory footprint of our solution is measured. The results show that it is possible to gradually introduce Java in 800xA using LJRT, which is the main contribution of this thesis.

Estimating CVD Thickness through Statistical Inference Methods

by Gian Antonio Susto

In this work we illustrate a Virtual Metrology system for Chemical Vapor Deposition (CVD) processes: on the basis of... more In this work we illustrate a Virtual Metrology system for Chemical Vapor Deposition (CVD) processes: on the basis of the available metrology results and of the knowledge, for every wafer, of equipment FDC variables, the system predicts CVD thickness.
Various modeling techniques are proposed, among them, Back-Propagation Neural Network is shown to achieve the best performances in terms of prediction accuracy.
Due to the high number of physical parameters involved in the CVD process, variable selection techniques have been used; on the other hand, due to the huge fragmentation of data derived from CVD’s mixed production, statistical data clustering have been adopted to increase the number of homogenous observations.

A Predictive Maintenance System for Epitaxy Process

by Gian Antonio Susto

Co-authored with Alessandro Beghi, Cristina De Luca, Michael Holzinger, Martin Huber

Process temperature has a strong influence on the growth characteristics of Epitaxy processes. We propose a... more Process temperature has a strong influence on the growth characteristics of Epitaxy processes. We propose a Predictive
Maintenance System able to estimate wafer temperature behaviour.
The algorithm is based on a Markov Chain Model and can handle the presence of different recipes run in the equipment. The proposed system allows the operator to estimate the probability that a control action on temperature is needed in a finite number of process iterations.
As a second outcome, the algorithm provides indications on the magnitude of the control action to be performed. The system also includes a module for removing outliers and one for change-points detection to guarantee robustness of the prediction in case of drift due to maintenance events not registered by the equipment.