Evolving Excellence

The Nimitz Goes To Home Depot

By Kevin Meyer

Regular readers know that I have a particular disdain for overly complex MRP and ERP systems, and recommend spending about 0.1% of the cost and going to Staples instead.  Simple visual solutions using whiteboards now control the shop floors in several companies, even very large ones.

Thanks to Jason for pointing me to an article by Andy Sernovitz at Damn I Wish I'd Thought of That confirming I'm not totally nuts.  Andy took a tour of one of the most leading edge pieces of technology... very very complex technology... an aircraft carrier.  Filled with high tech systems in every nook and cranny, but guess how they manage their flight deck: with a visual board.  I do need to apologize to Andy for quoting so much of his original post, but it's important to my point.

The entire flight deck is organized by a simple table-top diagram called the Ouija Board. It looks like this:

This is an incredibly important job. His job is to make sure every plane is in the right place, fueled, and ready to take off, while making sure the runway is clear, and no one gets run over, during war, at sea.  Imagine trying to park 60 semi-trailers on a 4-way freeway intersection, during a pre-school soccer game, in a thunderstorm, during a Harley rally, and never block traffic.  Except they're airplanes, on a boat, full of bombs. Look closely and you'll see that he uses metal models, washers, bolts, and thumbtacks.  Why?  Because it never breaks. 

Now I know what you're probably thinking... in the realm of $500 toilet seats those toy planes probably cost a couple hundred grand, right?  Nope.

He told us he was offered a multi-million-dollar touch-screen computer system to replace it.  His answer?: How do I fix it in the Persian Gulf under attack, or during a typhoon?  I've got everything I need for $10 from Home Depot.  And a spare set in a tackle-box under my desk.

That's not the only example.  How do you think they track the thousands of bombs moving all over this monstrous ship?  Yep, you guessed it: whiteboards.

The location of every bomb on the ship is tracked, in real time, on these whiteboards.  Thousand of bombs have to move across dozens of decks and find the plane to get loaded.  You only have minutes to get it right.

So the next time someone tells you you have to have a complex software system to manage your shop floor, remind them of the Nimitz.  Spend some time leaning your processes, reduce excess material and wasteful activities, and then utilize the most simple... and inherently intuitive and robust... visual management method possible.

A Finger for ERP

No I didn't give ERP the finger.  Mike Wroblewski did via an article in Reliable Plant.  As regular readers know, I'm not exactly a fan of big overly complex software to manage production operations that are often inherently simple.  And every time I spout off on ERP systems, often referencing one of our most popular posts on using a whiteboard and some beer instead of software, I get a few more comments saying that ERP is inevitable.  Luckily Mike and his company, Batesville Casket, believe otherwise.

Anyone of us in business today that has to deal with enterprise resource planning (ERP) systems knows that while these systems are meant to make life simple for us, it more times than not works in the opposite direction. We are constantly battling to schedule our work to match our customer requirements based on the output of our computer system. To borrow a quote by Japanese NUMMI leaders as reported on Curious Cat blog, “computerized inventory systems lie”. Regardless of all our efforts to keep current and accurate information fed into the mouth of the computer ERP system beast, our computer system never matches the dynamic, real (physical) world. In my experience, it’s true that computerized inventory systems lie.

Mike then discusses how his company is using an alternative, with some inspiration from, ahem... well, you know.

It is just part of our lean focus in making our parts flow using simple visual means instead of using ERP solutions within the plant. We even have started using the whiteboard scheduling approach in our dry lumber storage before cutting. The whiteboard approach has been recommended many times in the past by Kevin Meyer at Evolving Excellence. The goal is to make it simple and visual.

Batesville Casket isn't exactly a small company.  Neither are many of the other companies I know that use manual, visual shop floor scheduling.

How about you?  Are you continually fighting inventory and scheduling glitches from your ERP system?  Perhaps it's time to focus on how simple manufacturing really can be and invest in a whiteboard.

Waste and SAP?

I appreciate Bill stepping up to fill the Evolving Excellence blogosphere over the past couple weeks while I deal with some family medical issues.  Also great timing as no one is better than Bill at insightful comments on the auto bailout (err... sorry... "taxpayer partnership") quagmire.

It's been many months since I've picked on SAP and bloated ERP systems in general.  Most readers know my thoughts on the issue... from suggesting a $100 whiteboard instead of a quarter million minimum on SAP, or the chasm of ERP complexity.  A few days ago I came across an article that was just too much to resist commenting on.  Not necessarily because of the primary premise... alleged misrepresentation in SAP's product demos... but the fact that the customer pounding on SAP is Waste Management.

There's just some sort of beautiful irony in a company called Waste Management sueing SAP.

But the irony doesn't stop there... the product demo in question has gone missing.  Isn't it the purpose of a garbage company to make things "go missing"... I bet they're experts at that.  Although Waste Management accuses of SAP of similar magic.

Ok, presumably this has whetted your appetite a bit and you want to read some of the juicy story.

A presale product demonstration software package that is key to Waste Management's rancorous lawsuit against SAP has gone missing and both sides are claiming the other should have it, according to documents filed in a U.S. District Court.

The trash hauler has said SAP used "rigged and manipulated" demonstrations during sales presentations. A motion Waste Management filed May 18 said the demos were a key element of the "false representations" SAP made to "induce Waste Management into entering a software licensing and implementation agreement."

Waste Management has been asking SAP for a copy of the demo since mid-2008, according to the motion. But in a May 22 filing, SAP said it had the system until August or September 2006, but does not any longer.

SAP has "searched extensively" for the system and wants it "as much or more" as Waste Management, since it "will help SAP disprove WM's fraud claim," the filing states. In addition, SAP said it has produced ample discovery materials related to the demo, such as its scripts and hundreds of e-mails.

In fact, Waste Management should have the demo in its possession, as it was transferred to the trash hauler's system in late 2005 and early 2006, according to SAP, which demanded in a May 14 filing that Waste Management turn it over.

Moral of the story?  Be careful of product demos.  Might be a good idea to keep a copy of the demo to see if the final installation provides the same functionality.

And if you "transfer to [a] trash hauler's system" you might want to consider that they are experts at making things disappear.

Thoughts on Communication Poka Yoke from Japan

Coincidentally while I was winging my way to Japan a few days ago, fellow lean blogger Mark Graban wrote an interesting piece on technology and poka yoke in the context of aircraft.

This reminded me of something I read on the way home. USA Today (via a WSJ Blog) reported that the recent airliner crash in Spain was caused by the pilots forgetting to set the flaps properly for takeoff. This was human error. But.... the technology failed them. The warning system (a "poka yoke" of sorts) failed to sound an alarm before takeoff, leading to the crash. The USA Today reports that this same error (forgetting to set the flaps) has occurred 50 times in the US recently... but the warning system worked.

So, sadly, it's just a matter of time before the same crash is repeated.

Of course that made me feel good, being on a lot of planes and such this week.  While on approach into Narita I came across another example where a good poka yoke is needed: tower communication.  The geek in me likes to listen to the pilot/tower communication on channel 9 on United, and this time I almost wish I hadn't listened in.  The tower guys were providing landing instructions in English, but with suvch a heavy Japanese accent that I had a very hard time understanding them.  I thought this probably wasn't that big of a deal since the pilots and tower guys were probably used to communicating this way, used a relatively small number of phrases, and knew the routes and landing patterns. 

Until I heard a Northwest pilot nearby have to ask the tower controller to repeat his instructions three times before it was understood.  And a minute later hearing an Aeromexico pilot trying to communicate in heavily Spanish-accented English with the heavily Japanese-accented English tower controller.

There's got to be a better, safer way.  If cops can communicate to squad cars using electronic messages to onboard laptops, you'd think high tech aircraft could too.  Generally I'm not a big fan of software and technology, and prefer simple simplicity, but this is probably a situation where a white board may not cut it.  Or garbled verbal communication for that matter.

Travel safe.

An IT Guy Gets Lean

Regular readers know that we often focus on how software solutions can create problems for lean.  This isn't to mean that all software is a problem, and we definitely believe that software tools, appropriately deployed, can create value.  But more often than not they mask waste and reduce the ability to change by enforcing rigid processes.  Once you reduce inventory and move to pull production, is MRP really necessary?  Does a complex system to manage NCMR's, CAPAs, and other exception documentation really add value, or should the focus be on reducing root causes that are creating the need for those documents in the first place?  You get the picture.

Apparently some in the IT world are starting to understand.

IT a roadblock to progress? Never, you say. How can that be when IT lives and breathes innovation? Who else but IT dared to usher open source, XML, SOA, and cloud computing out of the high-tech labs and into production systems? Nevertheless, there is one aspect of technology where all of that daring doesn't seem to be in play.

Could he... really... be talking about lean?  Yep.

Manufacturing has traditionally been a push system. Up until lean, business analysts would forecast demand, and manufacturers would ship products to market based on those forecasts. Unsold goods would sit in the warehouse or distribution center either waiting to get into the pipeline or never to be sold due to a missed forecast.

More than a decade ago, Toyota created kanban, a system that builds to demand rather than forecast. It is a pull system based on what the market currently needs. Demand doesn't have to be just from a downstream customer; that is not the only "market" recognized by Kanban. If a product ships from a distribution center, the system recognizes that the center needs more of that product, as Kanban is continually sending pull signals to an upstream work center.

When done right, inventory can go from being held for a month down to a day, or it might never even see the inside of a warehouse, going instead from the manufacturing plant straight to the customer.

Nice summary from an IT guy!  So how does that rock his world?

SAP or Oracle MRP are a problem because they cannot set up an "execution" system to perform based on lean principals. SAP and Oracle are irrelevant; both are good planning tools, he says, but planning is batched-based and does not involve real-time execution. Customers typically trim back on their ERP systems. It is still the system of record, but they tend to turn off the MRP function.

"One of the things MRP does is to take a forecast into a consideration of how we order," says John Young, materials and supply chain leader at Trane Residential Systems, a manufacturer of HVAC systems and subsidiary of Ingersoll-Rand. "In a Kanban environment, we basically throw that away." Young says that he still pays attention to the forecast as a planning tool, but because of Kanban, everything is "replenishment"-based. In the last 120 days, the company has saved about $300,000 in parts held.

I'm sure SAP and Oracle would dispute that they are purely batch-based, as both have supposed "lean modules."  But we're admittedly biased so we'll let it ride.  The author then tries to simplify the concept further to help out his readers.

Young describes Kanban this way: It's as if you lived 100 miles from the closest grocery store, and if you need to buy a container of milk, you would buy two gallons at a time. But now with Kanban, you only need to buy one gallon. It is this new concept that is the biggest stumbling block for IT in terms of adopting lean manufacturing. Most major companies have invested multiple millions in their ERP system, and it's IT's job to run the system.

On top of that, these software-acquisition decisions for the major ERP systems are made by the CEO and CIO, who don't understand the shop floor.

Oh yes, how many times have we heard that directive from the top?  The IT groups need to step up and see how they can help create value for the company, even if it doesn't directly create work for their departments.

If IT views itself as a cost center that just says, "Yes, sir, we can do it," then nothing changes. If, however, IT decides it has something of value to contribute and suggests there is a better way, someone upstream might just listen.

Bottom line, manufacturers are struggling, and all IT is saying right now is, "Don't worry, we're going to put an Oracle system in place, and in three years, all your problems will be solved. Trouble is, the manufacturer may not be around in three years.

Think about it.

We have.  We're glad you are too.

IW Swings Both Ways on Robots

I was thumbing through the latest online issue of Industry Week and came across an article that raised my hackles: What Happened to the "Auto" in Automation?  Why?  Because IW, of all rags, was promoting robots.  Let's dive into it...

The headlines blared: "North American Robot Orders Declined 17% in First Quarter." Bad news? Yes, but the good side is that "non-automotive orders jumped 36%.

The good side may even be better, adds Rohit Khanolkar, engineering manager at Applied Manufacturing Technologies. His speculation hinges on whether the rest of manufacturing has come to realize how much the automotive industry gained from the lean flexibility of robotic automation. "The automotive industry has taken a lead role in selecting robots as their preferred flexible automation solution," Khanolkar says.

Yeah, right. Give me a second to get myself off the floor and stop laughing. The auto industry really is leading the world right now, isn't it?  Ok, there's also a little fuel cost thing skewing, or skewering, the industry a bit.  And remember that old Superbowl commercial where GM disrepects robots?  But let's move on before I start laughing again...

"Industries other than automotive need to realize the benefits that robots can add to their manufacturing processes. Robots are accurate, flexible, programmable, can work in environments not suited for humans, and more importantly, they are repeatable with a high degree of accuracy. These attributes directly contribute to end product quality, which in turn reduces scrap and rework -- a requirement of lean manufacturing."

And they can't think.  When was the last time a robot submitted an improvement suggestion?  Participated in a kaizen?

The article goes on to describe more of the glories of robotics, stats on how many bazillion have been deployed, etc.  I won't bore you with that; by now you know my opinion.

But then, since I felt I needed another good laugh, I clicked on a "related article" titled Seeking the Lean Potential of Robots.  I fully expected to be wowed by more robotic godliness, especially since the article was written by the same author as the first, John Teresko.  I was wrong. 

"Avoid the mistake of buying a robot and just being content with the expectation that lean process results will automatically occur." That's only the first caution from John Burg, president of Ellison Technologies Automation. His caveats continue: "Improve the task or process before automating it -- and get your organization to agree to the cultural shift that a lean commitment requires. Only then will robot tending of a machine tool or process pay off."

Holy cow... he gets it!  And believe it or not, I'm not necessarily against all robots; I'm against the danger of robots covering up or automating waste.

He recommends a formal investigation to validate the presence of lean processes. "Go beyond merely accepting verbal claims from the engineering department. Despite their assurances on lean, we often discover that machine operators are compensating for a whole host of inefficiencies in the cell.

"In at least 70% of our robot implementations, we find that production management may not be fully aware of the cell operating deficiencies that operators have to contend with."

Burg says, "The conventional goals are double-digit performance gains, otherwise nothing is done. And with robots, in an opportunistic sense, U.S. users tend to look at the automatons as a solution while they're actually only potential opportunities."

The other side of the coin is that a proper lean analysis may also show that there is no case for a robot implementation, adds Burg. "Correctly assessing the lean opportunity for robots is a significant consideration -- both for the customer and for us."

Remember, John Burg is the President of a robotics company.  And he's saying that "a proper lean analysis may show that there is not case for a robot implementation."  Hallelujah... there are a few good robotics guys out there!  At least one.  In fact, I'd suggest that if you are contemplating using robots, you should contact Mr. Burg to get an honest assessment.  It sounds like he'd have the guts to assess your lean implementation first.

The greatest strategic advantage, Burg adds, is to use lean practices as a competitive tool to maintain U.S. manufacturing -- to save the facility from outsourcing.

What a breath of fresh air.

SAP Gets Fat, You Get Thin

So you've taken the dive into SAP or Oracle instead of simply investing in a few whiteboards and some beer.  That's too bad, but it's not too late.  You could still try the simpler approach.  Or stick with SAP... and now realize you have to pay even more on top of the original software, licenses, maintenance fees, and implementation wizards. Yes, SAP wants more out of your wallet.

Two of the biggest makers of business software have raised prices, a sign that consolidation in the industry may be easing the competition over prices that has been a hallmark of the last decade.  SAP AG said Wednesday that it has raised prices for the ongoing support and customer-service fees, known as maintenance, that are part of large software purchases. The move comes one month after rival Oracle Corp. raised list prices for its software.

I'm sure there's a reason, right?  Nope.  It's just because they can.

Unlike price increases for food, fuel and many other commodities, the changes in software don't stem from a shortage of supply or a rise in demand.  They are attempts by software makers to increase their bottom lines, said Brendan Barnicle, an analyst at Pacific Crest Securities Inc.

And that increase can hurt other projects.  Projects that can directly impact competitiveness.

Owens-Illinois is in the middle of installing SAP's software, and Mr. Masney said he is compensating for the cost of the project by cutting spending in other areas.  Mr. Masney also buys database software from Oracle and said the new price models for both companies could have a "ripple effect" for tech departments.

So what was the cost-benefit analysis on SAP again?  And how does that compare with some investment in making the your operations lean and efficient, plus perhaps a couple whiteboards?

The Squirrels of SAP

Thanks to fellow blogger Mike for pointing me in the direction of one of the best SAP pieces I've ever read, yes even including mine.  The author begins by trying to understand a particular peculiarity of squirrels.

You are driving down a street in your car and up ahead there is a squirrel at the side of the road eating a nut. You aren't on an intercept course, there is no way you are going to hit that squirrel. So what does the squirrel do? At the very last possible moment, rather than watching you drive by, THE SQUIRREL DARTS STRAIGHT FOR YOUR CAR, passing inches in front of or behind the front tires.

Why does he do that?

After considerable consternation and testing of hypotheses, he comes to a conclusion.

Sure, my car is bigger and faster, but the squirrel is smaller and quicker, with a heart that beats up to 700 times per minute. To the squirrel I seem to be driving by in slow motion, and whether he goes in front of the tires or behind or in front of one and behind anotheris strictly a matter of style: once the squirrel has my vector, Victor, he's in command.

The squirrel doesn't trust me. Sure, it looks like I'm not even chasing him, but he's a tasty squirrel and I'm a saber-toothed tiger. By waiting until the last possible moment then running TOWARD me, the squirrel is rushing the net, moving the confrontation effectively forward in time in such a way that the squirrel is pushing his tactical advantage.

As a predator, I'm simply not supposed to expect this squirrel to be running toward me, rather than away. He's using the element of surprise to confuse me. And it works, because I've never hit a squirrel with my car.

I know I know, what the heck does this have to do with anything, let alone SAP?

SAP and companies like it do something similar by making powerful software that is quite deliberately difficult to use. They could make it easier. Heck, the capability to make it easier is shipped right with the software, though never pointed out to the customer. I used to think this was a matter of geek machismo, where higher value was placed on processes that were more difficult to command simply because it could be used to maintain for the techies an upper hand against management. But now I think it's much simpler than that and SAP just wants its software to be more difficult to use because that maximizes revenue. It is more nuts for the squirrel.

But of course.  He then takes an enlightened view of ERP systems in general, one those of us who have dumped ERP shop floor controls for a white board understand all too well.

Sometimes ERP systems come about as a response to inadequate IT, but more often it is just a very expensive alternative to walking around and talking to employees. Putting in an ERP system isn't going to improve the business by itself: you still have to figure out what the data means and make decisions.

Implementing a big ERP system -- any ERP system -- is expensive. The problem is there is not enough return on investment from the ERP system itself to justify the cost. You need more. The real savings must come from improving your firm's business processes.

Which of course drives the need for wizards.

That's why there are so many SAP consultants. And that's why SAP, itself, makes 40 percent of its revenue from providing consulting services -- revenue that would be significantly less if the software was easier to customize and easier to use.

If SAP software was easier to customize and use, SAP the company might get a few more customers but would have significantly less revenue. Or that's the fear.

Robert goes on to describe some of the technical nuances of SAP that make that possible, including embedded capability that the customer isn't told about.  Pretty ingenious... I guess.

Lean ERP Survey...

I know, I know.  Most of you already have a pretty good feeling for how I feel about ERP systems and lean manufacturing.  But this time I owe someone a favor.  A bright young guy that interned with me a year or two ago is running a survey for his masters thesis on Lean and ERP. If you are interested in participating in the survey click here. The user/pw is "user" and "leanerp". You can get a copy of the results, and we'll be sure to comment on them here.  Believe me...

Feeble Attempt, No Cigar

We haven't heard much from the False Gods lately.  Perhaps our rapid identification and interdiction of their new battle front forced them into retreat.  Earlier this month they dared poke their squirrelly heads above the sand, so I guess we better take a pot shot out of sheer boredom.  Let's first take a look at their cowardly attack.

Let's come clean - lean manufacturing isn''t living up to everyone''s expectations. While some well-known companies such as Toyota have attributed success to their lean manufacturing approach, others are still struggling to realise the improvements they expected.  A survey from management consulting firm Bain & Company finds that just 19 per cent of companies that have tried lean manufacturing principles are happy with their results. So what is standing in the way of better returns?

Ah yes, the same old Bain study.  It is accurate, but is being used by every nickel consultant to peddle a variety of "solutions."  Being a minion of a false god, what does this particular author believe the problem is?

The key concepts surrounding lean manufacturing were defined 40 years ago and today''s manufacturers inevitably face new trends and challenges. One new trend is mixed-mode manufacturing whereby elements of build-to-order, engineer-to-order, and assemble-to-order can be applied to support an ever-increasing product portfolio and fluctuating demand profile. This model is more complex than the simple make-to-replenish model, and Lean practitioners are struggling to adapt lean principles to it.

We're far more complex, faster, global... you get the picture.  So obviously we need...

In today's complex manufacturing environment, lean manufacturers need IT solutions...

But of course.  Not.

As we've mentioned time and time again, the real reason for most lean failures is that companies don't recognize that there are two pillars to lean.  The first is the waste reduction, value creation, etc etc that can be effectively implemented with a variety of tools.  The second is "respect for people."  Understanding the value of the knowledge, creativity, ideas, experience, and commitment of employees.  Even when it doesn't show up on a P&L or balance sheet.  Most companies, even ones committed to lean, don't understand that concept.  Those are the companies that tout their lean initiatives while at the same time laying off thousands of workers just to rehire them in some cheaper location.

Let's briefly get back to the proposed IT solutions, just for grins.

...lean manufacturers need IT solutions that can: Handle hybrid make-to-order fulfilment models using a single technology framework; support additional material control techniques beyond traditional kanban control; use a single application platform capable of addressing lean requirements during design, operate, sustain and improve phases; standardise on a technology platform that can handle plant-to-plant variations; provide a closed-loop environment to drive structured kaizen improvements using real-time metrics; and maintain the same level of simplicity, user control and visibility for which lean is known, while making the enterprise rollout of a lean programme possible.

And be forced to do it the same way over and over, per the methods and processes determined by the software manufacturer.  What would happen if you came up with an improvement, a change, that conflicted with those set parameters?  Open the checkbook.

Instead of embedding unchangeable complexity into software, you could learn from some real lean companies, even very large companies, that rely more on simple visual controls and simplified processes... and especially the power of people.

Manufacturing really isn't all that difficult and complex.  We just make it that way.