Resources for Global Manufacturing Excellence

This newsletter is public.  Forwarding to colleagues is encouraged.

Sponsor: 

Agilonics - agile manufacturing for new technologies.  Oddball spares, prototypes, retrofits, and obsolete product getting in the way of your regular product?  Let Agilonics take care of it!  We have experience with electronics, photonics, medical devices, and consumer electronics.   More info.

Contents

1.  Happy New Year from Superfactory!
2.  Update on the Job Shop Lean project
3.  Letter to the Editor - Theory of Delays
4.  Toolbox and Lean Series downloads
5.  Featured Book: "Implementing World Class Manufacturing"
6.  Article: "Lean and Flexible: A Way Forward for High Variety, Low Volume (HVLV) Environments"

1.  Happy New Year from Superfactory!

Best wishes for a prosperous 2002!  The past year was a difficult one for many of us, but leading-edge organizations have taken advantage of this opportunity to improve their systems and capabilities in order to capitalize on the next upswing.  They will also capitalize on less innovative companies that applied traditional short-term thought to trim capacity and costs without improving methods.

New markets and new paradigms will become important.  More than ever, lean, agile, flexible, and the power of knowledge have become critical when creating excellence.

2.  Update on Job Shop Lean project

Superfactory is a co-sponsor of the Job Shop Lean project, a collaboration created by Dr. Shahrukh Irani of The Ohio State University.  This four month old project has been a large success, and the associated discussion group has over 320 professionals worldwide investigating the peculiarities of lean manufacturing in high-mix low-volume environments. 

There are now 17 free downloadable files on the JSLEAN board, ranging from performance measures and metrics to articles on applying finite capacity scheduling to job shop environments.

Visit the Job Shop Lean project.

3.  Letter to the Editor - Theory of Delays

The past two Superfactory newsletters have included both parts of the "Theory of Delays" by Vincent Bozzone.  As expected and intended, this somewhat different perspective has generated considerable healthy discussion, which helps all of us analyze and learn about new concepts.  Superfactory welcomes comments about content and ideas.

Bill Dettmer of Goal Systems International wrote us with the following detailed analysis of Theory of Delays.  If you missed the original articles, part 1 is in the October newsletter and part 2 is in the December newsletter.

Now... from Bill:

I just read "The Theory of Delays (TOD) - Part II, Improving Performance and Profitability in Job Shops & Custom Manufacturing Environments," by Vincent Bozzone of Delta Dynamics, Inc.

In the middle of the article, Mr. Bozzone says:

"3. A delay can be caused by a physical constraint in the production system (TOC), but not all delays are the result of constraints. The large majority of delays result between process steps, not within a step."

The "but" in these statements would (erroneously) lead readers to conclude that TOC is only about physical contraints in production processes---meaning the physical processing actions themselves. This would seem to indicate that Mr. Bozzone has read no more about TOC than THE GOAL (Goldratt, 1986), and probably misinterpreted some of what he read in that book. The second sentence, above, confirms this conclusion. Anyone who has read anything about TOC written in the last 10 years would know that it aims to identify and break system-level constraints. If the constraint of the whole system is a particular production process, then Mr. Bozzone's comments would be accurate. But where TOC is concerned, the actual process activity might represent the system constraint only 10% of the time for most systems---if that often.

Near the end of the article, Mr. Bozzone says:

"Measuring Results: In addition to broadening the system scope and focusing attention on delays, TOD uses two operational measures that are meaningful. One is similar to that used by TOC—throughput productivity which is measured in sales dollars per payroll hour. The problem with using this metric by itself is that customer service can be sacrificed for throughput efficiency.  Therefore, it is also necessary to measure lead time (and on-time ship performance) to ensure improvement is happening in these central areas as well."

This is an exceptionally misleading paragraph. Besides trying to link his operational measure with TOC, he implies that TOC uses this metric alone.  That's patently false. A thorough reading (and understanding) of the TOC literature would reveal that TOC does NOT suggest sacrificing customer satisfaction for throughput efficiency. In fact, it does exactly the opposite. For years, Goldratt has been trying to get the business world to wake up to the fact that striving for internal efficiency is counter-productive to achieving a company's goal (which is presumed to be increased profitability).

By the way, for those who might be interested, the real (i.e., accurate) TOC definition of throughput is sales revenue minus totally variable costs (very close to what traditional management accounting calls "contribution margin"). The metric Mr. Bozzone cited isn't even close to a commonly-used TOC metric, which is "Throughput per unit of the constraint's time," NOT "payroll time." The difference is crucial. Sales dollars (as used by Mr. Bozzone) is revenue alone. It ignores the variable cost of each product.  The time referred to is actual clock time. The "payroll" part of his definition has no meaning.

Other than these two misrepresentations (probably mistakes resulting from incomplete research), a substantial part of this article (not all, but most) constitutes the TOC philosophy, "hijacked" and dressed up with another name ("Theory of Delays"). And the outcome isn't even as good as the original.
Who was it that said, "All things old are new again"?

I would encourage readers not to take my word for it, however. Judge for yourselves. There are several books that explain the Theory of Constraints quite effectively. Any one of them will substantiate what I've said here.  A few pertinent ones are:

* Goldratt, Eliyahu M. THE HAYSTACK SYNDROME (North River Press) - for the unvarnished truth on throughput and "throughput per unit of constraint time."

* Schragenheim, Eli. MANAGEMENT DILEMMAS (CRC St. Lucie Press) - for the unvarnished truth on what TOC is really about (optimizing whole systems, not just production processes)

* McMullen, Thomas B. THE THEORY OF CONSTRAINTS MANAGEMENT SYSTEM (CRC St. Lucie Press) - for a different perspective on what TOC is all about.

BILL DETTMER
 

4.  Toolbox and Lean Series downloads

Superfactory's downloadable Toolbox and Lean Series continue to be very popular.  Several new files have been added to the Toolbox, and an update will be going out later this month to everyone who has previously downloaded the Toolbox and subscribed to the update mailing list.

The Lean Series, a group of very concise presentations on the various facets of lean manufacturing, has received a 98% customer satisfaction rating by users.  We are making some modifications to the presentations to improve on that number!

Visit the download page.

5.  Featured Book

Implementing World Class Manufacturing - A Shop Floor Manual

This is a "hands-on, how-to" guide to making significant improvements in a manufacturing plant. Uniquely written in a step-by-step format for both managers and operators, Implementing World Class Manufacturing allows everyone in a facility to participate in making the business the best at what it does - now and in the future.  Read more.

Visit the Superfactory Library.

6.  Article

The following article is one of the many available in the public domain on the JSLEAN discussion group.

Lean and Flexible: A Way Forward for High Variety, Low Volume (HVLV) Environments

Dave Williams LLB (Hons) MSc
BAE Systems Avionics Ltd.
Email: dave.williams2@baesystems.com

Introduction

In this paper I intend to set out a proposed way forward for using Lean Thinking outside of the Low Variety, High Volume (LVHV) environment.  My own background is within a Spares and Repairs unit supporting the Aerospace Industry. This type of business is typical of a High Variety Low Volume environment.  We are expected to support a wide range of products with perhaps only one or two orders for many of the products occurring each month.   My proposals would also apply to jobshop environments where the customer is looking for manufacture to order (MTO) capability.

Problems with the two dominant approaches for implementing Lean Thinking - Lean Manufacturing and Theory of Constraints - have been identified as follows.

1.        Clear parallels with High Volume Manufacture cannot be drawn when it comes to the target of inventory reduction because there isn’t a high level of repeatability of orders. Vincent Bozzone comments “Lean does not fit custom manufacturing very well because the focus of the methodology is on reducing inventory. This is not a problem in a make to order business where no finished goods inventories exist (by definition).  Lean works best in mass production manufacturing facilities that build finished goods inventories to stock.  This is where the money is... a $1,000,000 reduction in inventory saves the company the "cost to carry" which often runs 25% or more per year, so there are big savings in inventory when lean programs are successfully implemented.”

2.        At first appearance, it does not seem possible to identify a specific “constraint” as with TOC.  I again quote from Vincent Bozzone’s article …. “TOC, on the other hand, focuses on eliminating constraints that impede the flow of work through the shop (throughput).  The problem here is that constraints or bottlenecks are moving targets in a jobshop environment.  Bottlenecks are created and disappear hourly with changes in customer priorities and schedules.  An operator fails to show up for work.  Result?  A bottleneck is created.  A customer changes a specification.  Result? An existing bottleneck may disappear, and a new one may or may not be created. The progress of an order through a shop often has any number of starts and stops that have nothing to do with constraints.  The Drum-Buffer-Rope technique, central to the Theory Of Constraints, is not particularly useful in the dynamic world of the jobshop vs. a more stable high volume production environment where a constraint is more likely to stay put”.

I believe that it is possible to find a way forward (by) drawing on the concepts and the thinking approach inherent within both Lean Manufacturing and Theory of Constraints.  Our approach should be to use the logic without just jumping in and grabbing particular solutions.  In other words, there are some principles such as the 5 Lean Steps that are key to Lean Thinking.  We should be starting with questions about these.  What we shouldn’t do is rush in and assume that Lean means “reducing inventory.” Or that constraints are “resource problems” and that we must therefore implement Drum-Buffer-Rope (pull) scheduling.

Some things that I think a jobshop environment can and should be looking at are:

1.       Identify your key Value Stream

Lean Thinking tells us to map the value stream for each of our product families.  People in jobshop-type environments say that it is too complex to map these.

Peter Keen in “The Process Edge” argues that the reason that a lot of companies fail even though they have demonstrably improved their processes is that they improve the wrong ones.  He uses a tool called the Salience/Worth Matrix.  This looks at:

1.        Salience: how prominent is a process

2.        Worth: this identifies if the process is an asset or a liability

There are several types of processes:

1.        Identity Processes: The process that identifies you viz. the reason that a Customer comes to you (to do business with) and not someone else

2.        Priority Processes: Those processes you need to do in order to make the Identity Processes happen

3.        Background Processes: General day-to-day background ex. admin, time collection, etc.

4.        Mandatory Processes: Processes you have to do because of the law or various regulations. There are a lot of these in my industry!

There are also Folklore Processes – these are the ones you believe you have to do because you always have done them.  In fact, there is no need or requirement for them. They may have been mandatory at some time in the past but no longer.  You just scrap these!

Traditionally, if you were taking a Lean Manufacturing approach, you would identify all your different product families and then you would map the value stream for each. You may have a different product family for every day of the week; so, in the HVLV environment, that approach will not work.

However, there is a particular reason why your customer comes to you as opposed to either Joe Bloggs down the road or a particular manufacturer.  For example, an Airline will send something back to Boeing or Smith’s Repair Shop because they are the OEM and have design authority/specialist knowledge A lot of energy is wasted trying to gain competitive edge in areas where the Customer expects that element of the service to be a-given.  For example, take your typical Aviation Repair facility.  We could put together teams to find out ways of reducing the time the Customer has to wait for receive back his repair(ed) kit.  We manage to reduce his waiting time from 90 days to 80 days.  We are then very pleased with ourselves.  We have beaten the competition, they are static at 85 days!  Then we are surprised when our Customer doesn’t get over-excited about our achievement.  Why is this?  Well, the Customer really wants his kit back immediately.  So, taking a couple of days off the  Turn Round Time (TRT) is neither here nor there.  We are working towards achieving something that he thinks we should have achieved anyway.  So why should he give us any brownie points for that? I need to concentrate on improving the processes that give me competitive edge.

So this is our first clue.  We know exactly what our Customer is looking for from us.  We have a Value Stream.  We can then map all the steps required to please our Customer and deliver the unique service he is looking for.

2.       Constraints in a Service Environment

The next Step we should tackle is to identify how we can improve our competitive edge by focusing on our Identity Processes.  This is where Goldratt and TOC enter the picture! Why do you have a problem when an operator doesn't show up?  Why does a Customer changing his spec cause a major headache?  Surely the aim of the jobshop is to be as flexible as possible.  In an ideal world anybody would pick up the job and run with it.

It is unlikely that the constraint is a specific operation (remember Goldratt argues that things like investment, market, etc. can be constraints).  It may be something to do with how your Human Resources department operates! 

I would suggest that Knowledge Management is often the biggest Constraint. If I knew what you knew, then you would have less of a problem.

Let’s take a case study.  Someone has given the example of not being able to produce quotes on time. I'm very familiar with that problem. Why can’t I get the quote out on time?  Answer:  Because I need input from somebody with specialised knowledge and experience in estimating to research the quote.  Then I will need somebody with specialised commercial knowledge to approve it.  Unfortunately, those people come in small quantities and are either over-worked, on holiday or ill! One solution is to authorize somebody to be able to take the Request for Quote and see if it matches criteria for a previous quote.  We can also look to fix/band price our activities (people pay for the service not the product).  That means that you can get the quote out quickly and not have to wait for the specialist to be there or somebody to approve it.  Only a small percentage will go through the expert.

3.       The Cost of Flexibility

If my views are right, then a key tool that we are going to need is one that shows the Cost of Flexibility, just as TQM requires us to know the Cost of Quality.  I expect this to look something like the types of models used to determine Economic Batch Sizes.  A company would be able to show its Economic Flexibility Rate.  This will help the Customer to see that he can get the response times he wants but that they come at a price ex. we can do all his repairs within 24 hours but we will have to flood our process with people and hold a large stock of components.  The next stage will be to do what the Japanese have done with economic batch sizes and start to break down the barriers to quick setup changeover and batch sizes that conform with actual order quantities.  We will have to think about how we can change the process and challenge the assumptions so that we can have total flexibility without impeding on costs (This is the Service equivalent of a batch size of one which, from a manufacturing point of view, means total flexibility).

4.       Knowledge Management 5S

Earlier I talked about Knowledge Management as the key.  I talked about looking at the information flow as the Value Stream.  If we should be thinking of this environment as service first and manufacturing second, then knowledge is the true product.  It is why the Customer comes to us.  In this case, Information is as important as inventory.  So, one of the things that you need to be thinking about is;  How do we ensure that the information is accurate and easily available?  A 5S approach helps to achieve this.  You do not want any redundant or wrong data lying around!  You want to know where to go to get information and to be able to trust it when you get it.  This is the same as the requirements for tools and inventory in manufacturing.  So a process like 5S that gets rid of the crap and sorts everything so that there is a place for everything and everything is in its place will help.  It means making sure that Business Systems are kept up to date and don’t need a massive data-cleanse every two years.

5.    Summary

If I were forced to give a fancy title to my approach, I would call it Identity Process Management.  A company that knows what identifies itself as being a unique provider of products or services and understands what it needs to do to be competitive.

So what to do?

1.        Know your identity.  Don’t put together a huge Customer Survey so that your customer can tell you what you already know, i.e. that you are always late and that your Help Desk staff are rude, etc etc.  Just ask him one question, “All things considered, why do you keep coming back to me”?

2.        Put as much investment as you can into making that one thing the best of its kind.  Make sure that you have a reputation for it.

3.        Make sure that your other processes are aligned to ensure that your Identity Process always performs

Question:  Does this mean that I neglect my other processes?  Do I accept that I will always have long lead times?  Answer: No!  But only do something if you are going to make a real difference.  When you have got TRT down to five days, then you can continuously improve it to get it down to 1 day but at the moment you want to take big chunks out.  Secondly, if it isn’t an identity process but is still important to the customer, then be prepared to pool resources with partners and even competitors to negate the problems with that process.

None of this is rocket science.  It’s basic commonsense.  The only obstacle to making these step changes is our willingness to make the change.

References:

Superfactory is always interested in your comments, suggestions, and submissions.  Please contact us!

© 2002 Superfactory.  All rights reserved.