An Ah-Ha Moment With A Simple PDSA Flowchart

Recently I saw Karen Martin post a PDSA flowchart on Twitter.  The graphic does a wonderful job of outlining the steps of problem solving and PDSA thinking, highlights the importance of defining your problem and how much time you might need to spend before moving forward in the process, and covers the question of “what does adjust really mean?”.  There’s a good chance printing this out and handing it to my leadership team in our next capability building session will generate some rich refresher discussion.

Karen Martin PDSA Flowchart


But there was one line on there that really sparked something in me as I read through it.  Line 3 says “Set a Target Condition”.  Seems simple, seems in line with most problem solving training… so why was I so interested?  What sent my brain off racing to analyze this line?  This line really resonated with me because many recent kaizen events I’ve been involved in struggle and bog down at the point of discussion about the target, especially if they are trying to go somewhere they’ve never been before vs. getting back to a level they’ve obtained sometime in the past.

Yes, the problem is the difference between target and actual, and understanding the target condition is critical – you have to know where you need to be in order to understand why you aren’t there.  But this is not always easy, especially in fledgling lean organizations.  I believe manufacturing environments in a company attempting to go through a lean transformation are able to understand and articulate a target condition much faster than those in the office or product development organization.  That’s still not saying it’s easy!  But they tend to have more historical data, understanding of their manufacturing process and where defects occur, where their problem modules are and where to focus, and how quality, cost, and production rate have trended.  Many times you can go back several months or even years in the data and say, “oh, we clearly have trended downwards from where we once were”, or “oh, we clearly made an acute shift in June, you can see it right here on the timeline”.  It’s much easier to do this step when you have relevant data available, and when you have people who know how to mine it.

But take for example people in an organization dedicated to innovation, who are attempting to do a kaizen event.  Many times the most difficult question is “what should the target be?” because you don’t have a database telling you what the health of your innovation process is, or how many defects that process produced over the past 3 months… or even a clear definition of what a defect is!  The same holds true for most office processes:  there simply is not a plethora of data to sift through in order to identify targets and potential causes.  So in these types of kaizens, either as prework or as a part of the kaizen itself, sometimes the group must develop a system to measure what is going on in order to even understand if they really have a problem!

So back to step 3 – “Set a target condition”.  I just like the meaning of the process behind this so much better than something along the lines of “What is the target condition?”  Asking “what is it” implies that there IS one out there in your organization and you either don’t know it or can’t find it.  The concept of “setting” a target condition seems to empower the group to define what that target should be:  either because they know what it should be, or because they decide to set a target that corresponds with some level of improvement from where they think they are.  And if they are able to do that, they are able to move to the next step!

I haven’t put it into practice yet, but I believe sending that strong message of empowerment to a kaizen team will enable them to align much more quickly in an event, and move forward to defining causes and countermeasures rather than arguing over what the target actually is.  Because no matter what – you’re in that room because something was not up to snuff, and the bottom line is that is doesn’t even matter what the target is – all that matters is that you are not yet where you need to be.

See a great guest post by Karen on Mark Graban’s blog about PDSA vs. PDCA, and another spirited post by Mark himself on the subject.

Do You Need Advanced Math Skills To Be A Scientist?

Today I was doing some quick surfing to catch up on some of the latest headlines.  I don’t quite remember how I got to the story, but I found myself reading a Harvardvery long and detailed article on a member of the Harvard Quiz Bowl team accused of cheating, subtitled “the biggest scandal in quiz bowl history.”  Reading through the article, I was surprised at how the story spilled on to three pages, and impressed by the level of research and detail put into an area that many across the country may dismiss as trivial and uninteresting.  Then, I reminded myself that I spend several hours a month writing about lean thinking…

The article was on Slate.com, and I found it to be written with much higher quality and deeper insight than say, the “Top 10 Grocery Store Traps” on the regular MSN homepage.  So I tried another one, entitled “E.O. Wilson is wrong about math and science“.  EO WilsonThis Slate article examines a recent WSJ article written by E.O. Wilson, an “eminent Harvard biologist and best-selling author”, and concludes that Wilson is telling aspiring scientists that they don’t necessarily need mathematics to survive – and takes issue with that message.  (It was a much shorter read than the aforementioned “scandal that brought down a powerhouse”)  The article itself didn’t strike a specific chord with me, but then I began to read over the comments by readers.  Most of the respondents appeared to be physicists, professors, scientists.  In general they argued against the Slate article’s conclusions about Wilson’s message, with several stating that although understanding of basic math is necessary, most advanced math concepts are not necessarily applied in many scientific fields, and that Wilson was saying that “discoveries can come from ideas, not always just number crunching”.

Now I’m certainly not here to argue one viewpoint or the other.  Reading through the comments and arguments, I felt woefully inadequate on the intellectual front.  However, the articles and comments made me think about principles of Lean, a little on how Six Sigma fits in, and what it means to have GEeveryone be problem solvers vs. a handful of experts.  I was trained in Six Sigma at G.E. back in 2000 (thanks Cindy!), and loved the tools, the logical thought process, and the focus on reducing defects.  It gave you a roadmap of how to analyze and solve problems, especially when dealing with multiple interactive variables.  Then in 2008 at my current company, I learned about lean, and saw how much more power it held for the entire workforce, the entire end-to-end system, the entire organization, than Six Sigma alone did.

Lean is about creating a community of scientists, from floor operators on up through senior leadership.  It teaches that you do not need experts to come in and solve all the problems, scientistbut that to truly be a continuously improving organization everyone needs to be able to solve problems everyday.  To enable this ideal world where everyone solves problems, we were not told to go teach everyone triple integrals, how to design a fold-over DOE, or how to complete a one-way ANOVA test.  We were told to teach everyone the scientific method, typically summed up in Lean teaching as Plan-Do-Study-Act/Adjust – hypothesis thinking.

“I understand my problem to be this.  I think the problem is caused by this.  Let’s check and see if that’s really the cause.  If we do this, then we expect this… Yes it is, ok let’s figure out how to block/eliminate that cause and make sure we never have that problem again.  If we do this, then we expect this… And let’s figure out how to make sure that our solution stays in place.”

I think just about everyone can understand that basic line of thinking.  If you assume 1-5% of all the work you do in your organization is waste in the eyes of the customer, and you want to strive to eliminate it all… well, that’s a lot of Six Sigma projects, if that’s your only improvement tool.  And you’re going to want an army of blackbelts.  But you don’t need to build that army!  I can’t remember exactly where I read it, but I liked someone’s recent description of using Six Sigma to solve all your problems – it was something along the lines of using six sigma for everything is like using a flamethrower to remove small stacks of hay in your yard.  So much of that waste could be removed with much simpler methods by so many more people if you train them in the fundamentals.

Is there a place for Six Sigma?  Sure!  It is a great tool.  Like I said, I loved using it.  There are complex problems with multivariable interactions that need to be analyzed with statistics to help make the right problems visible and weed out the noise.  But it will still take hypothesis thinking to figure out what you’re going to do to solve those problems!

Think of your organization and all the work you execute as a Louisville Sluggerlarge, round, 4-foot diameter chunk of lumber.  Think of what your customer values within that chunk of lumber as only the shape of a baseball bat.  Lean thinking, tools, and principles, are what can help whittle away large chunks of wood and get it to the shape of that baseball bat.  Six Sigma might be a tool you can use to fine tune where the sweet spot is when you get to that point.

One question I don’t have any answer to yet is, when do you decide what you do everyday isn’t enough and call in the experts?  When is the proper time to pull that particular andon cord?  What defines a problem as “complex”?  I’m sure there are some simple decision rules, such as “We’ve tried four times and the problem keeps coming back” or ” we’ve tried 10 experiments and don’t have any more direct causes we can think of to check”.  But my hypothesis is that most organizations have a long way to go and a lot of benefits to realize after they begin their lean transformation before they even need to begin worrying about that final 5% of waste to eliminate.

So do you need advanced math skills to be a scientist in your organization?  I think you need the ability and discipline to form and study a hypothesis.  And I think anyone can be taught those skills, regardless of whether they took AP calc or not.

I Came, I Saw, I Problem Solved… I Only Got To A Temporary Countermeasure. Do I Still Feel Worthy Of Lean?

Problem solving to root cause is an important skill to build capability in.  The “why” behind problem solving to root cause is so that you prevent problems from ever recurring again, thus eliminating potential waste from your future PDCAactivities.  Along the way you grasp the situation, take steps to contain the problem so you are still able to provide the product or service to your customers, understand potential direct causes, test the connection from cause to the problem, develop and test a countermeasure (your hypothesis towards preventing the problem from ever coming back), and put measures in place to sustain and check that the countermeasure works and is still in place.  (Yes, this is simply a description of P-D-C(S)-A with a containment step thrown in!)  It doesn’t matter if you’re talking about 4-Step, 8-Step, x-Step or DMAIC problem solving, the fundamental principles are all the same.

Most organizations that begin a lean transformation are already very good at what they think of as problem solving.  A problem comes up, I work hard to understand what caused it, I fix what caused it, and we’re up and running again.  I add it to our troubleshooting guides and therefore if it happens again we will be able to get up and running even faster because we know how to fix it!  And the veteran problem solvers will be able to tell you war stories of all-nighters where hours of investigation finally yielded something they never thought of: a motor wired backwards and turning in the wrong direction, a bit set wrong in the control logic, or an incorrect part sent and installed that looked just like another part but had different guts.  They might call it “stuff that never should have happened if someone else had done their job right in the first place”, because they are still learning what it means to focus on the process and not the people.  And how if only they’d called the expert in the first place they could have avoided all those hours of downtime because he’d recognize the symptoms and connect it to a problem he solved 5 years ago and could have told them exactly what to check and fix.

“Haven’t we worked on this problem before?” and “Didn’t we fix this last year?” are common phrases you might hear that should trigger you to wonder if you really understood the root cause of the problem the first time.  It feels like problem solving deja vu!  You honestly shouldn’t feel bad about it though, especially if you are still at the outset (read: first several years or perhaps decades depending on point or systemic problems!) of your lean journey.  Solving to root cause, so the problem never occurs again anywhere in your organization, is hard.  It can be hard to identify the root cause, hard to rollout countermeasures across groups in multiple global locations, hard to not get distracted by all the other fires you need to fight this week that seem like a much higher priority.

I recently switched cable, internet, and phone service providers from Time Warner to att uverseAT&T, for a whole host of reasons that could be turned into another post on “thinking customer”.  I’m actually very happy so far in a short period of time with what I now have from AT&T Uverse.  It’s not all ice cream and puppy dogs through the switch however, I have had a couple of internet setup issues that have been frustrating, but they’ve actually been more due to lack of knowledge and added system complexity on my end vs. something that was the service provider’s responsibility.

Yesterday I woke up to a new problem.  When I tried to open a web browser, I got an error message from my AT&T wireless router that popped up on the screen, saying “Excessive Sessions Detected.”  It explained that one of my computers had a whole lot of internet sessions going at once, and that it was likely the result of some form of a virus, or malware.  So, my head began doing problem solving.  Target = Able to access the web from all my devices.  Actual = Not able to from one PC.  Let’s continue to grasp the situation.  Check PC #2 – I get the same error.  Check IPhone – I get the same error.  Actual now equals “Not able to access web from any devices.”

Excessive Sessions Warning

Potential direct causes… 1) the error message tells me it may be a virus 2) the error message tells me I may have gaming software causing it 3) could be a problem with the router 4) a power problem or connection problem somewhere in the system 5) something is broken on AT&T’s end 6) my internet cache is full 7) just something weird that requires a restart.

Ok, so let’s try and work through the most likely causes – 2) I can eliminate this cause, don’t have any gaming software going on (sad, I know, I’m a long way from my college and single days!).  4) check to see I have power everywhere – all ok, eliminated.  1) the system error is telling me “virus” is the first place I should look.  So I run a check, and sure enough, it finds two items and eliminates them.  So as I restart my computer my mind is already jumping down the why chain to root causes like, inadequate standard for setting up my antivirus software, or inadequate process for selecting antivirus software.

Computer is restarted, and… nope, error message still pops up.  How about 7) – let’s restart the router and everything else.  Nope, error still there, can cross that one off.  Now it is about time for work, so I eat some breakfast, watch some TV, and off I go.  Midway through the day my wife tells me the TVs no longer work.  So now, new information surfaced that tells me that something in the system is degrading – the problem is getting worse!  So in my head I mapped out how the system worked and where problems could occur (see setup picture below), and couldn’t figure out what was changing to cause the new problems, because most of the TVs don’t run off the router.  Why did they work in the morning and then suddenly not work in the middle of the day?  Now I start leaning towards something on AT&T’s end as the direct cause.

ATT Setup Map

When I got home I hopped on the phone with AT&T, explained the situation and what I’d done so far, and then we went through their troubleshooting guides.  We do a reset from their end, restart computers and routers and DVRs, and the error still comes up. We cleared the internet cache and tried again.  Still have the error.  We’ve now eliminated 5) and 6), and AT&T is out of ideas on their end too.  Their only solution left is to send out a technician tomorrow, and maybe they’ll swap out a router to try and check 4) – the only direct cause we have left on the list.  This disappoints us, because we wanted to watch the new Modern Family!

As AT&T is finalizing the order for the technician, I decide to check one more thing.  On the error message there are two buttons – one says “Do Not Show” and the other says “Continue”.  I had tried “Continue” early on and didn’t get anywhere.  “Do Not Show” was labeled as something you should only click if you think the cause was gaming software 2) which we had already eliminated, and I didn’t want to ignore the error if I really had a problem or a virus.  So at this point I said what the heck and clicked “Do Not Show”.  It asked me to log in to my gateway, I did, and then it gave me a message – “The problem has been resolved.”

docmcstuffinsEureka!  We were now connected to the outside world again!  My three year old would not be without her Doc McStuffins in the morning!  We could watch Modern Family!  I could stream YouTube videos through my TV again!  I was the hero, I had “resolved the problem”!

My lean thinking was gnawing at me though.  I didn’t know what caused the problem in the first place.  I can’t recreate it.  I can’t develop any countermeasures to prevent it from happening again.  And I’m not sure the AT&T person really captured my “solution” in their knowledge database so that they try it with other customers in the future before deciding to send out a technician.  Yes, the problem is contained, and we are up and running again.  Is that good enough for this situation?  Or should I have done more?

Good organizations are very quick to recognize and contain problems, and to get up and running again to avoid customer service issues.  Great organizations have the discipline to spend time working towards truly understanding the root cause of the problem, developing adequate countermeasures, and ensuring waste never recurs.