The Golden Bee-Bee

Most understand what we mean when we say “critical path”.  To help us keep focused and in context for this blog entry, the critical path describes the longest path in duration that needs to take place, which in tern becomes the minimum time line of a project.  To expand the thought a little further, also think about it as a bottleneck in a process causing it to be the critical path.

Lean manufacturing, Kaizen, TPS, look for incremental improvements.  It does not really matter what part of the process you apply it to, because your intent is to eliminate waste of all forms and in any process.  Most of this is accomplished through observation, or perhaps observation and data.  This simplistic concept of incrementally eliminating waste is to sustain the improvement once it is proven and accepted.  It takes effort from the team or individual who observed and defined the problem, who gave it proper scope and planning, successfully implementing (the most difficult part), confirming the problem is corrected with any supporting tools in place, making it a Standard.  This is the PDSA (plan-do-study-act) cycle (or PDCA for check if you still prefer).

A LinkedIn post (which linked to a presentation) caught my attention and frustrated me when I read it.  I jumped to a conclusion that someone was preaching once again “This Is The Way”.  I am not too far off in that, and even I can be that way at times in my own preaching.  What the reader (meaning you and I) need to remember is the person writing it, whether the post I read or what I write myself, most likely has a target point to make.  One that is specific in example and solution, that indeed will work every time or almost every time.  Whether we convey it properly in clear thought or writing can be another story.

The presentation I am referring to is titled, “Let’s Retire the PDCA Wedge; What really keeps performance from slipping back?” by Mike Rother and Jeff Uitenbroek, August 2011.  The intent (and correct me if I am wrong when you look at the presentation) is that something is throwing a wrench in the works, keeping us from moving forward with continuous improvement.  The thought is our definition of a Standard may be the cause.  It was also mentioned that no matter how well a Standard is implemented and maintained it will go through entropy (degrade and become a waste of energy).

The presentation then sited some information on Deming’s PDSA, suggesting that the cycle needs to keep turning for never ending improvements.  The presentation also mentions Toyota and the difference between Standard and Standardized work.  Standard meaning something you want to achieve, and Standardized work as operating as specified by the Standard.  It also implies that in “wedge”¹ thinking, when we slip backwards we believe we lack discipline and want to blame someone, whereas the Toyota way of thinking recognizes an abnormality that we just have not figured out yet.  Sort of like a Kanban being an admittance you do not know how to go 1-piece flow yet.

For me the Deming cycle is about continual improvement.  However Deming's focus is generally on reducing process variation.  There is not a guarantee you will produce quality.  That is dependent upon the how the process was setup and maintained (another form of Standard).  There is never a focus on the team member who is willing and able, because we know he cannot affect the output of the process.  The process is the process.  However continually improving will guarantee uniformity (good or bad) at a low cost.

Toyota has gone to great lengths to specify what they do is Standardized work, and not Standard work.  I have gone through Toyota’s supplier training where this was drilled into me.  Toyota defines Standardized work as, “. . . organizes all jobs around human motion and creates an efficient production sequence without any “Muda”. Work organized in such a way is called standardized work. It consists of three elements: Takt-Time, Working Sequence, and Standard In-Process Stock.” They went out of their way in class to discourage the use of the word Standard.

I have seen the “wedge” graphic with and without the wedge, and really never put any emphasis on it or used it in my education of others.  I do know that when QS-9000 converted over to TS-16949, there was an emphasis to change “continuous improvement” to “continual improvement” (as stated by the AIAG personnel who educated me on TS).  The purpose was to get away from the mind set of continuously improving without checking that you sustained first.  The graphic was a straight line ramp for continuous improvement vs. a stair step process of continual improvement. 

The “wedge” presentation did suggest that one reason we use this mindset is to comply with audits.  I would rather hope the audit system changed their process to meet what we need to be doing in our process.  The presentation also suggests we do need a Standard in order to satisfy the customer, however the Standard itself will not stop entropy.  So the suggestion is, the Standard needs to be a target, a set of conditions to be met in order to satisfy the customer.  This is what others and I call “positive tension”.  You put enough spring force (improvement) pulling on the PDSA wheel to keep it rolling at a given velocity (so that you do not become cyclical), resulting in it going up the incline indicating continual improvement.

I do agree with the “wedge” presentation that there is never “steady state”, but “constant change”.  Whether we like it or not this is very true.  However, it is also human nature to not want change, or to view change as a bad thing.  To me this is because life, for the most part, moves slowly for us.  We do not see the changes per se because it creeps up on us.  This gives us time to absorb, condition, and make it a norm or Standard in our lives.  Business is different.  In the never ending quest to bring in more revenue, to increase margins, we choose to want to do better than the Jones' as it were.  If they have it, we want it and we want something better, all to maintain current and to attract new customers.  Change is quick, requires thinking, requires revised processes (continual improvement), requires us to work, looking at an ever changing target, rather than taking life slow in increments to yesterday’s Standard.

I do contend that maintaining a Standard will keep the process from slipping back. (Similar to the stair step graphic for TS-16949.  Once sustained on the step you do not roll back.)  The only entropy taking place should be when the Standard becomes outdated through Kaizen.  The high complexity of business makes this so variable that we cannot draw a line in the sand and say “removing the wedge is the way”, or even “using the wedge is the way”.  You need to see Your problems and solve Your problems.  Can the “wedge” presentation be Your solution?  Sure.  Can it be the wrong solution?  I am just as sure.  One application I believe the “wedge” presentation will always work is in marketing.  You will need to continually innovate to not only satisfy, but also delight the customer.  Where it will not work is automotive Tier 1 supply.  Having a Standard (which I should really say "Having Standardized Work") helps maintain a statistically controlled process, and enables you to problem solve when the process becomes out of control and to Kaizen.

I also want to point out you cannot just continually improve willy-nilly.  If you do not improve on the critical path, you will not increase the velocity of the process, which is generally where our problems lie.  (Will using the Standard as a target necessarily put us on the critical path?)  If you incrementally improve, eliminating waste, there is a cost savings and some increase in velocity.  What should happen as you continually improve along the non-critical path is that the “bottleneck” in the critical path will expose itself, indicating your next continual improvement event.  As you improve you will have a shift in the bottleneck within the processes, and you attack that next.  Is the bottleneck always seen?  No.  Can the bottleneck be calculated?  Yes, if you have all the variables.  Observation is generally easier because of all the variables.  VSM (value stream mapping) also helps to find and see all the variables.

Using the “wedge” can help you determine if you worked on the bottleneck in the process or not.  What is your goal, your target?  If you stopped to observe the new Standard, has the system improved?  Whatever the Standard is, plateau or target condition, it should have removed the problem, satisfying and attracting customers through a system that is aligned to the customer, and through processes that support the system.

So, “This Is The Way” will never work across the board.  However you can “Choose Your Path” to make it work.  Be specific in your quest to solve a problem.  Do not look for the Golden Bee-Bee.

¹ The “wedge” thinking is the PDSA wheel going up an incline, using a wedge to keep it from rolling back.  As you continually improve and move forward, Standards are put in place as a wedge to keep the PDSA wheel from rolling backwards.

Who's the Customer? Are your priorities set right?

Okay, I’ll say it again, “Who’s the Customer?”  Should I repeat it?  It is a simple thought.  Did I get you to drill down to whomever consumes the product / service / or information?  I may have.  If not, then good, because there are many customers to satisfy, with the primary one being the end consumer.

In this blog entry I want to emphasize the support needed from management to provide the product / service / or information to the consumer through VA (value added) activities.  Along with the VA there are many NVA (non-value added) activities.  Then there is the in between, the gray area, the “incidental work”.  It is NVA but essential in conducting the business.  I am not going to define each here because that is not the purpose of this blog entry.  I will, however, defer you to this link should you have questions:  “Where to Begin with Lean: A3 Analysis” by James Womack

Value is added at the contact point of the product / service / or information.  As the contact point proceeds through the process, it carries the same level of importance at each VA activity.  When that activity is upset in any way, it now becomes the focus, to resolve whatever issues that made it go awry.  This is where the accountability board needs to be placed to log the issue and to assign a time and date, along with who is accountable, to permanently correct it.  Supervisor Gemba walks need to frequent these boards (along with shift pre-start meetings at these boards) to understand missed timing and to get things on track through problem solving countermeasures.  The department manager needs to daily Gemba walk these boards (and initialing), making sure there are no barriers to resolution and to verify accountability (that the countermeasure is in process or completed).  The department manager also enhances any countermeasures or ads to the supervisor countermeasures, along with answering to upper management as to how this happened.  Upper management needs to Gemba walk these boards weekly to assure adequate resources were given, accountability has been met, and that permanent corrective action is in place.  This is what industry terms "leader standardized work".

Let us refocus from leader standardized work back onto the customer and priorities. Look at just the VA activities from the beginning to the end of the process.  The people performing these activities are your internal customers.  Can you prioritize which of these customers is the most important?  If you have a normal, stable process, one that is on track, the answer is “no”.  If something has upset the VA activity anywhere along the value stream, then that activity or customer takes priority.

The point I want to make is, the further away from the VA activity, with respect to your own activities, the more you become “support” to the internal customer that needs help.  Many times people will say they need to finish a report or take care of a broken sink (something to that affect).  How is that aligned to the True North of customer satisfaction?  So what are your priorities?  It should be any VA activity that needs help, at any time.  We are all service to that end.  Once the process is again stable, then we can re-direct some of our energy to the incidental work, and never to the waste unless we intend to eliminate it.

Management needs to step back and observe, asking “are we satisfying the system or the customer?”  Too many times I see we have layered on processes or reasons to do things for the sake of the system, and the customer ends up suffering whether they are internal or external.  The sad part is, we never see the suffering of the internal customer because we maintain our silos first, rather than aligning our activities to the internal customer needs, which intern are aligned to the value stream of needs including that of the end user.  This is why people preach that you need to understand your process, inputs, outputs, and to identify bottlenecks in the process.  This is where we VA, which ends up adding waste if problems are not addressed.

So I will ask again.  Who’s the customer?  Are your priorities set right?

Whether to Weather the Storm, or Peel Back and See

Running a business is basically the same, whether big or small, whether product / service / or information.  You have a process with inputs and outputs.  Multiple processes are linked through a value stream if you will.  The value stream extends from the raw material suppliers like mining iron or growing cotton, to the end user of this product / service / or information.  The question that needs to be asked is, “how aligned to True North is your value stream to the customer?”

How smooth we communicate those hand offs from inputs, through the process, to outputs converted to inputs to the next process, is what Lean manufacturing is all about.  It does not matter what size, color, religion, or political view your product / service / or information is.  What matters is how you communicate the inputs you need, so you can process, and provide the correct output to the customer.  You tell you suppliers what you want, when you want it, how much of it you want, and at what price.  You process the input by adding value.  The product / service / or information is transformed to the output to the customer.  This can be one process, several processes, and with as many inputs and outputs as needed.

What is the point here?  The point is, no matter how you look at business, you are satisfying the needs, wants, and desires of the customer through a thought out process that has inputs and outputs.  Do you have an established process?  Have you defined the output requirement to the customer (internal or external)?  Have you defined the inputs to the process that will give you the needed output?

I know if I were to provide the product / service / or information just by myself, that I could provide the correct output from my process(s) given the correct inputs, and in the correct sequence of events that need to take place in the process, with less waste than an entire company would.  Problem being, it would take me forever compared to an entire company.  So our need is to take that single person value stream thought and convert it to a multi-person task through the value stream.  Making sure we well define the process, outputs, and inputs.

Realize I have steered you to the thought of process, outputs, and inputs.  This is typical Six Sigma practice: knowing what your process is capable of, then what is the process to output to satisfy the customer, and finally what inputs do you need to do that.  What breaks that mold is when a process does not exist to provide the product / service / or information.  Then you would start with the output and work backward through the concurrent design of process and inputs.

The concept is simple.  Having everyone buy into the execution seems to be what creates the storm.  The silo mentality of “I need this first before I will let you do that”, which typically does not fully align to True North to satisfy the customer, gets in the way of our performance.  This causes gusts of delays.  This causes the missed “pitch” or “lead” if you will, making someone think that they can squeeze in a different project while they wait.  This intern begins the roller coaster of more unintended missed pitches in the process, turning the process into a hurricane.  Hence my leading paragraph into these blogs that we need take the risk of peeling back the layers of work around and to do it right the first time. 

Is it a bad thing if someone waits?  Can you not visually see them waiting?  Would you not ask yourself “why are they waiting?” and do something about the process?  It seems like waiting is a good thing if you can see problems as they occur.  The trick now becomes managing solutions to those problems as quick as possible, incrementally correcting the system and its processes. If the person instead of waiting filled in the time with another project, you would have never known you had a problem until it was too late. It is the worker that is being driven by old school supervision, creating an environment that “if I look busy I won’t get yelled at”, or in the supervisor’s case “if they look busy everything must be working fine.”

Whether to weather the storm, or stop and peel back the layers of work around, seems to be the decision.  Storms within your process, or incoming to the process, will create waste with a great potential of defects or defective products / services / or information to the customer.  As you allow the storm to take you away from True North, the competitor who can better steer True North will always make it there first, in a manner of speaking.

Quality Control - Before or after?

A LinkedIn forum question asked for a laymen’s explanation as to why QC Inspection is not a way to ensure accuracy of the end product.  As usual, responses from others were lengthy and not necessarily directed at answering the question.  This includes one I had posted approximately 30 days earlier, when I must have been upset with some of the responses.

If a defective product was made, inspecting it at the end of the process does not make the product any less defective.  So now it has to be scrapped (or if a simple defect reworked).  Who is paying for this?  Is this not one of the forms of waste?  Is this not what we are suppose to be continually improving, the elimination of waste? When waste is created, in any form, the true value is lost and the customer ends up paying for the value + waste through higher selling prices that cover the accepted projected loss of the manufacturer.  This is what opens entry to market, giving competition room to squeeze between you and the customer, eventually taking your business away.

The purpose is to make it right the first time.  Quality at the source and the responsibility for that quality comes from the person adding the value which is the operator.  And I must emphasis here, WITHIN THE PROCESS THE OPERATOR WAS GIVEN. If your process has been setup to manufacture the product to customer specification, and the process is statistically in control, then there is no need to inspect the end product right?  The process is the process, good or bad.  The process data is what "speaks" back to the operator indicating whether the process is in statistical control or not, and the operator is to react when it is not in control, eliminating the chance of defects.

Now, if you are saying defects will still occur in a statistically controlled process, then so be it.  That is what the process was designed to do.  You need to change the process to reduce (eliminate if you can) the defects.  You need to identify the defect and root cause.  You may need to reengineer the product for manufacturability so defects cannot be made.  You may need to revisit the PFMEA and add a poka-yoke on a value added feature the operator has difficulty maintaining.  You may need to revise the standardized work to add quality check points.  There are many things you can do, however realize it comes from the process and that is what needs to change.  Please understand what you want to do is to change the system or it's processes to aid the operator in creating the value and reducing (if not eliminating) waste.

As I read the aforementioned forum question and responded as best I could in laymen terms (there is always room for improvement), I thought of problems with processes I have encountered in automotive.  I would estimate 60 to 75% of the process issues were caused by variance from incoming material, which the process was not designed to handle.  As I usually do I asked Why? maybe 100 times (okay, perhaps a bit of an exaggeration, however subconsciously I have drilled down through a lot of Why’s and I am zeroing in on root cause before I consciously think of the second or third Why?).

As I thought through this I realized, during product development companies hand make parts, do one off samples, 200 or 300% inspect prior to shipment, basically producing the perfect part.  Even through the different build events special qualified personnel are making the product.  Eventually the product is made off production tooling, still then by specially trained production personnel for pre-builds and line fill.  When mass production hits all of the “special handling” goes away, and you merge into the “true process” from not only your process but also from your suppliers.

What is supposed to protect the customer after product launch is the PPAP (production part approval process).  This is a warrant that the supplier process is capable of providing what the customer is paying for, whether it is volume of product or any fit/form/function of the product as specified by the customer.  How does this tie back to material creating havoc with the process and the potential of defects at the end of your production line?

Discussing this with my son Nick, one possible cause might be that during the initial trials and preparation for PPAP, not all of the supplier variations in the process were uncovered.  Perhaps the FMEA was not detailed enough.  Perhaps the supplier did not understand the process well enough to even contemplate a potential failure to put in the FMEA.  Perhaps the supplier kept making the "perfect part".  Whatever the case, when variation occurs that upsets the process, it must be identified and eliminated.  The intent should always be to provide materials that maintain a stable, statistically controlled downstream process.  If during mass production something is uncovered, fix it.  Don’t accept from the supplier, “Well it was PPAP’d that way and you accepted it.”  Go back and review the process, the FMEA, and update it with the new information so changes to the process take place that accommodate “the intent” of the PPAP, which is to make a defect free product from the process.

Quality control does not take place at the end of the line, nor at the beginning whether incoming material or something else.  Quality control is a constant effort in maintaining a statistically controlled process, identifying variance that was not originally considered in the process, and eliminating process variance that creates defects or defective products.

Okay, now let us get back to reality.  We all know quality inspections at the end or any other part of the process are not going away.  Why is that?  Quality inspections at the end of the process is just an admittance you do not understand the process nor if it is in statistical control.  This is no different than using kanbans.  A kanban is an admittance you do not know how to go to once piece flow yet.  You know one piece flow is what you need to do.  You just do not know how to get there yet.  So the purpose of all this writing becomes a "strive" for no inspections at the beginning or end of the process.  A "strive" for one piece flow, a "strive" toward continual improvement.  If you do not set the bar for "no end of line inspection", "no kanbans", then you will always fall short and the next company that figures it out will be more than happy to close your business for you.