Subscriber Discussion Forum:
FabTime Newsletter, Volume 27, No 3

As described in this month’s community announcements, you can use the new AskJen 3.0 chat engine in FabTime to ask factory productivity or cycle time improvement questions. The results normally draw on past FabTime newsletters, software tips, and help files. However, a member of our User Group asked an excellent question about including site-specific content in the responses. This is possible today, by adding site-specific help content to FabTime. AskJen 3.0 gives preference to focused tips (checklists), so we recommend these as a starting point. To add content, work with your FabTime support engineer.

What we’re sharing here is not a question that was asked by any individual subscriber, but rather, a resource that we think that subscribers to this newsletter may find useful. There are four books that we reference and recommend that can help those looking to better understand and improve wafer fab performance. For the first two, we (Jennifer Robinson and Frank Chance) wrote reviews back in the early days of FabTime. The other two we have not formally reviewed but do recommend. To see other books that we read and recommended during our FabTime days, download this PDF: Business Book Reviews.

Factory Physics, Third Edition by Wallace J. Hopp and Mark L. Spearman, Waveland Press, 2011.

Factory Physics is the kind of textbook that people refer to while doing jobs out in the real world. It places a mathematical framework around the behavior of factories, describing the underlying relationships with clear, easy-to-read examples and summing them up into straightforward Factory Physics Laws.

The book is targeted toward factory operations. The first section is a review of the traditional production management curriculum, including inventory models, MRP and JIT. The second section contains what we find the most useful material in the book, with chapters on basic factory dynamics, variability, and pull production systems. The third section puts the principles from the second section to use, and includes chapters on total quality management, production planning, shop floor control, scheduling, capacity planning and inventory management.

This book is filled with simple numeric examples and charts but does not shrink from including queueing formulas where they are necessary. The Factory Physics Laws are easy to remember, but also important. For example, Law 11 states:

“Pay me now or pay me later: If you cannot pay for variability reduction, you will pay in one or more of the following ways:

1. Long cycle times and high WIP levels.
2. Wasted capacity (low utilization of resources)
3. Lost throughput “

We recommend this book for those who have some engineering background but want to better understand factory operations and planning systems. It’s the text that we would use if we were teaching a course on the subject.

The Goal: 40th Anniversary Edition by Eliyahu Goldratt and Jeff Cox, North River Press, 2012.  Also now available as a 2024 graphic novel illustrated by Dean Motter.

This book has been widely read by semiconductor manufacturing personnel since it was first published in 1984. It accurately describes the behavior of manufacturing facilities, including such fundamental concepts as bottlenecks, constraints, and the impact of variability. One reason why it has been so broadly read is that it frames these concepts in the guise of a novel. This makes the ideas easy to read and digest.

The premise is that Alex, a factory manager, is given an ultimatum -- dramatically improve the performance of his factory in three months, or the facility will be shut down. Believing that traditional improvement strategies will never make enough difference in such a short time, Alex must resort to more desperate measures. He tracks down an old professor, now working as a consultant, and begs for advice. The advice of this consultant, Jonah, sets Alex and his team on a journey. Instead of just giving them the answers, Jonah asks them questions, and refuses to give more help until each question has been answered. As Alex learns through this process, so does the reader.

Some of the lessons of the book include the following.

• When you are productive you accomplish something in terms of your goals. Every action that brings a company closer to its goal is productive. The goal of a manufacturing organization is to make money.
• Because of variability, a factory cannot be run at 100% of capacity. Or, as Jonah says, “the closer you come to a balanced plant, the closer you come to bankruptcy.”
• One of the biggest problems in improving your factory is collecting the right data. Alex eventually concludes that “we're going to have to accept the fact that we're not going to have perfect data to work with.”
• “An hour lost at the bottleneck is an hour lost for the entire system … The actual cost of a bottleneck is the total expense of the system, divided by the number of hours the bottleneck produces.” This suggests managing bottlenecks very closely. This idea has spawned numerous consulting and software firms since the book was published.
• Non-bottlenecks do not need to be regulated so closely and should not be operated to maximize utilization. Jonah says that “activating a non-bottleneck to its maximum is an act of maximum stupidity.” [Side note: this is why striving for maximum OEE for non-bottleneck tools is unhelpful.]

We think that everyone who works in a manufacturing facility should read this book at least once. We re-read it at regular intervals and always find it insightful. Though we haven’t read it, there is now a graphic novel version available.

Optimizing Factory Performance: Cost-Effective Ways to Achieve Significant and Sustainable Improvement by James P. Ignizio, McGraw Hill, 2009.

While we have not formally reviewed this book, we did reference it recently when writing about The Waddington Effect (see Issue 25.06). Here is the most relevant part of the book’s description on Amazon:

“Focusing on the three primary enemies of factory performance―complexity, variability, and lackluster leadership―Optimizing Factory Performance cuts to the heart of the problem of less-than-world-class performance and demonstrates how those enemies manifest themselves in companies across manufacturing sectors. Ignizio also explores the insidious effect company politics and flagging commitment to manufacturing performance have on competitiveness.”

Note also that this book is an excellent resource for information about M-Ratio, which is discussed below.

Production Planning and Control for Semiconductor Wafer Fabrication Facilities by Lars Mönch, John W. Fowler and Scott J. Mason, Springer, 2013.

This text, published in 2013, bills itself as the first book to tackle the problem of production planning and control as applied to the complex environment of semiconductor wafer fabrication facilities. It covers the semiconductor manufacturing process, dispatching, scheduling, order release, capacity planning and production control (circa the early 2010s). The authors brought extensive academic and industry backgrounds to the project, and the book is full of useful illustrations and references. Recommended for industrial engineers and others looking to dive deep into these topics.

In the subscriber discussion forum of the previous issue (27.02), we discussed two metrics for encouraging preventive maintenance over more disruptive unscheduled downtime. We reviewed M-Ratio (a long-time metric used by people in the industry) and introduced a new, similar metric requested by one of our customers.

This is just a quick follow-up to let readers know that we have named the new metric PM-Ratio. We have added both M-Ratio (scheduled downtime % / unscheduled downtime %) and PM-Ratio (scheduled downtime % / (scheduled % + unscheduled downtime %)) to the data table for FabTime Tool State charts.

We also received feedback on these metrics from regular newsletter contributor Thomas Beeg, who writes the ever-useful Factory Physics and Automation blog. Thomas said: “my vote is to keep to M-Ratio, which is known to some extent already. Some of my thoughts on the topic I have shared on my blog here and here.” The second of these links includes the results of a poll on M-Ratio values achieved in the semiconductor industry (albeit one with a low number of responses), as well as proposed M-Ratio classes for wafer fabs.

While we take Thomas’ point that M-Ratio is already better known, we do think that there is value in making PM-Ratio also available. One nice thing about PM-Ratio is that it is reported on a scale of zero to one, whereas for M-Ratio there is theoretically no upper bound. To implement M-Ratio in FabTime, we had to select a maximum value for display, which we set to 10. Thomas’ survey results suggest that in the real world, M-Ratio values are highly unlikely to exceed 10, and in fact rarely exceed five. We also set M-Ratio to 10 when the denominator (unscheduled downtime %) is zero. We set PM-Ratio to 1 when the denominator is zero.

Why Track These Metrics?

• These metrics help distinguish planned maintenance (predictable, manageable) from reactive unscheduled downtime (disruptive, costly)
• On tool-state trend charts, these metrics reveal whether rising downtime reflects planned work or growing equipment problems
• On Pareto charts, tools with low M-Ratio or PM-Ratio are the highest priority for maintenance discipline improvement
• These metrics complement utilization metrics by revealing downtime quality, not just quantity
• Note: M-Ratio and PM-Ratio appear as data table columns only — they are not plotted on the tool state charts by default

Why Keep Both?

• Some sites use one…
• And some may prefer the other!

How do I display M-Ratio in FabTime?

To display M-Ratio or PM-Ratio by tool group in FabTime, use the “Edit Chart” functionality, as follows (the instructions are similar for display on a trend chart):

1. Create a Tool State Pareto chart. Slice by Process Family/Tool Group, set the date range to be at least a week, and filter to include your tool groups of interest (e.g. filter by module).
2. Click “Edit Chart” below the chart and/or data table.
3. Set the “On Chart” column for the “Avg WIP (wafers)” row to “Unused”. This is so that you can use the right-hand axis to display M-Ratio or PM-Ratio.
4. Click the box next to “Show All Fields.”
5. Find the row for either M-Ratio or PM-Ratio and change the “On Chart” column to “Bar” or “Line” as you prefer. Select a color in the next column. Change the value in the “Y Axis” column to “Right-Y”.
6. Change “Y2 Axis Title” to “M-Ratio” or “PM-Ratio” instead of WIP (Wafers).
7. Scroll up to just below the chart and click the box next to “Activate these edit-chart settings?”
8. Add the chart to a home page tab to save. An example of the resulting chart for PM-Ratio is below.
9. Because the scale of M-Ratio is ten times the scale of PM-Ratio, we recommend adding one or the other, but not both. If you would like to include both, one option is to remove all the other data from the chart (except for the “Process Family (description)” row, which is needed for x-axis labels), and put M-Ratio on one Y-axis and PM-Ratio on the other.

Past issues of the newsletter are available for subscribers to download in PDF format. Existing subscribers can find the archive link in your most recent email newsletter. New subscribers will see the link upon registering. 

• Register for the newsletter here. 
• Main article for Volume 27, No. 3: Ten Metrics for Fab Health
• FabTime Newsletter Community Announcements Page