Digital machine diagnostic boost for chip tools
Dutch research group TNO-ESI has brought all its diagnostics projects together under one roof.
This clustering at TNO-ESI in the Netherlands is paving the way for ASML and Canon Production Printing to benefit from each other in developing a digital assistant for machine diagnostics.
“In designing system diagnostics functionality, we’re used to following the Pareto principle: we attack the 20 percent of the causes that are responsible for 80 percent of the problems,” said Martijn van Veelen, system architect at ASML.
“Unfortunately, with this, we cover the large and mostly trivial part of the issues – those that occur frequently and that can relatively easily be correlated using data science technology. It doesn’t help us with the many rare cases that are similar but not the same. That’s what we want to tackle with TNO-ESI,” he said.
- Diagnostic tool eliminates blind spots in AI models
- Diagnostic simulation solution allows early testing
- Intel signs Siemens for digital twin fab digitalisation
TNO-ESI is working with the two leading chip lithography tool makers, ASML and Canon, on projects to provide more machine diagnostics to improve the time they are used in the fab and improve the overall performance. This is particularly important with the latest generation of extreme UV lithography systems that cost many millions of dollars.
“As of last year, we have a project cluster specifically focused on diagnostics. By bringing all our projects in this domain under one roof, we aim to stimulate knowledge sharing, not only between the teams but also between the partners, and get the most out of our combined efforts,” said Masoud Dorosti, director of science and operations at TNO-ESI
“This is a generic challenge with diagnosing complex machines,” said Peter Kruizinga, lead technologist at Canon Production Printing in the Netherlands. “When analyzing the different issues in the field, we also see a number of frequent problems alongside a very long tail of problems that crop up only very occasionally. Almost half of the issues encountered by our service technicians occur rarely. Together with TNO-ESI, we’re looking to develop a methodology for dealing with these types of problems.”
This includes developing reasoning models that produce hypotheses about what could be the cause of the symptoms being observing from the system. Unlike hardware failures, performance problems aren’t black and white, so the researchers are building a hypothesis generator for this gray area in machine diagnostics. From a mathematical point of view, this means moving from discrete random variables to continuous random variables, prompting the natural evolution from Bayesian networks with discrete states to a probabilistic programming approach..
After developing a machine diagnostics approach based on Bayesian networks, Canon shifted to addressing this system-level type of problems. It is currently mapping out the different failing mechanisms to see if they can generalize them and looking at probabilistic programming as a very promising research avenue.
If you enjoyed this article, you will like the following ones: don't miss them by subscribing to :
