Friday, June 05, 2009

Batch & Queue Construction?

Have you ever been in line at a sandwich shop where the "sandwich artist" was making three sandwiches at a time?  You know, he pulls out three loafs of bread, puts the meat & cheese on all three sandwiches, wraps each of them in deli paper, and hands all three to the clerk at the cash register.  It's happened to me, and I was not amused.  Why?  Because I was first in line out of the three, and I had to wait three times as long to get my sandwich.  Not only that, but I had mine toasted and it was darn near cold by the time I sat down to eat it!  Okay, I'm a little demanding of my eateries, but I can't help it; I'm a lean geek.

What does this have to do with Lean?  It's all about batch-and-queue vs. one-piece flow (click here for a good explanation from the Shmula blog).

The Case for One-Piece Flow

In the lean manufacturing world, the debate on batch-and-queue production vs. one-piece flow production has long since ended (batch-and-queue lost).  Anybody who has studied the Toyota Production System understands the superiority of one-piece flow: reduced inventory, early detection of defects, reduced transportation, more scheduling flexibility, and (most importantly) an increased organizational awareness of the need to solve problems preventively due to the lack of buffers in a single-piece flow.

So, where do we stand with respect to these principles in the construction industry?

Batch & Queue Construction

I've had a difficult time explaining my thoughts on this.  My belief is that in the construction industry, we typically exhibit the traits of a classic batch-and-queue operation, even though we're not technically batching buildings together.  Here are some of the similarities:
  1. Work is done by specialized groups (concrete finishers, framers, drywallers, plumbers, electricians, etc.) dedicated to performing a narrow scope of work
  2. These specialized groups try to optimize their own operations, even if that means a delay in the completion of a building
  3. Item #2 above means huge amounts of time are spent waiting for the product to transition from one specialized phase to the next
  4. Because of the delays between specialized work phases, the cause & effect trail of defects can go cold, resulting in fewer problems getting solved at the root cause
  5. There is little collaboration between specialized groups, resulting in less innovation
Hey, lean geeks, does this sound familiar at all?

Lean Construction

So, what is the lean approach to construction?  What could be done to move from a batch & queue approach to a one-piece flow approach?  Do the construction folks think it's even necessary to make this transition in order to achieve operational excellence? Your thoughts?  I certainly don't have all the answers, but here are a few pie-in-the-sky ideas:
  1. Create multi-functional work groups, consisting of highly-skilled installers capable of performing a wide range of installations
  2. Equip your highly-skilled installers with all the equipment (possibly custom-built equipment) necessary to complete a wide range of installations
  3. Utilize visual management to encourage participation through shared information
  4. Pre-fab as much as you can inside factories on lean assembly lines
I'm sure that experts like Dr. Michael Mullens with the Housing Constructability Lab or Hal Macomber at Reforming Project Management have a lot more to say about this subject than me, but those are my wild fantasies.  Are they feasible?  Your thoughts?


Matthew Horvat said...

There is massive opportunity with your pie in the sky ideas. But I think we have many lifetimes working on the fundamentals. Let's just focus on predictable work flow. Do that successfully and you will leave the competition behind. Leave the competition behind.

Michael Lombard said...

Matthew, thank you for your comment. I agree with your assessment; predictable work flow is a critical prerequisite to establishing one-piece flow.

What do you believe to be the biggest sources of process instability in the construction industry? My $.02: employee turnover, lack of visual standards in the workplace, excessive rework, and poor product design. I could go on all day!

Matt Stambaugh said...

You mention how different crews are most concerned with optimizing their own efforts, even at the expense of the larger project. I was just reading an article for my MBA class that talks about this problem of conflicting optimization among collaborative groups.

The argument was basically that if you can organize groups into just the right sizes and arrangements, this local optimization within groups will eventually produce a global optimum for the whole.

You know I'm a lean tyro, but does this make any sense within lean thinking?

Michael Lombard said...

Matt, do you have a link to that article? I'd like to read more about it to give a good response.

Just based on what you've described, I'm a little skeptical of the premise of the article. Yes, it's good that they are at least advocating that a global optimum is what's important, not the local optimums. But, trying to manipulate the size and makeup of groups to achieve better results sounds like the top-down, mechanistic, manage-by-results planning system that Sloan made famous at GM.

The problem with this approach is that there's no way for a management team to fully understand the complexities of system, especially when the circumstances change constantly. Therefore, it's impossible to plan properly for optimization at any level. The Law of Unintended Consequences takes over and you never get what you thought you'd get by manipulating the system.

It's better to create self-adapting teams that can sense changes immediately. If such a team is provided with enough shared information about the global optimization goals, they can tailor their local optimizations to the good of the entire system without the interference and manipulation of management.

Matt Stambaugh said...

The argument did strike me as being rather vague and provided no practical application advice. It was very much a theoretical article.

As for a link, the article is hosted on a paid academic hosting site, so a link would not be very useful. Below is the citation if you have access through your MBA program.

Kauffman, S. (1995, March). TECHNOLOGY AND EVOLUTION: ESCAPING THE RED QUEEN EFFECT. McKinsey Quarterly, Retrieved June 9, 2009, from Business Source Complete database.