[Off topic. I formatted all the links in this entry to open in a new window. Do you prefer it?]
For people unfamiliar with manufacturing, “mass production” is what any manufacturer does, spitting out carbon copy widgets on mind numbing production lines. Within manufacturing though, we know the definitions aren’t so simple. Just because a firm produces identical copies doesn’t mean they’re employing “mass production” ala Taylor, aka “push manufacturing. We know there’s a better way; lean manufacturing. Lean manufacturing means (in part) combining best practices of organized manufacturing systems with craft manufacturing, aka “hand made”. The latter by the way, technically means one person makes the entire product start to finish rather than through batch processing typical of “mass production”.
I’ve often thought that humans are genetically hardwired to batch processes. Take agriculture for example. You can’t pull a single serving of corn like you can with a coat or even a car like Toyota does. For much of human history, batching meant survival. That is not to say batching is efficient, survival usually isn’t. I’m guessing it’s hardwired into us because even children do it instinctively.
In the book Lean Thinking, the author’s daughter needs to help her mother by folding some flyers and placing them into envelopes. The author asks his daughter which method is better, to fold each sheet singly and set it into the envelope, or, to fold all the sheets in one batch and then set them into envelopes (another batch process). The child goes away and thinks for awhile and decides it’d take less time to batch the work. Anyone who’s read the book has thought of timing the process both ways but I doubt all but a few have. Fortunately for us, Ron Pereira has made a video to show the results using the example from the book. I strongly encourage you to see it; hopefully, it’ll change the way you think about making products.
I prefer the opening in a new tab. Thanks. Alan
That video is great! I’m glad he timed it out; -1 min from a 3.75 min process is fabulous. On another note, I much prefer the external links to open in a new window because sometimes I forget to open a new tab and it causes me slightly more difficulty to navigate back to this lovely site. However, I am lazy.
Q: I formatted… links… open in a new window. Do you prefer it?
A: Love it. If it doesn’t, sometimes I consequently get involved in a scavenger hunt.
RE one flow, now I have to streamline my dog’s food-making process. Kaizen project #2.
Interesting post. There are still times where “batch processing” is required in Lean. If there are areas of concern, in regards to being a bottleneck, then you will need to ensure you always have a plus quantity on the incoming side. This would be OK, and is not considered “mass production”.
The other reason for a single piece type flow, while it is faster, is also from a quality stand point. You catch your defects right away, and will only need to rework maybe 1-2 items, not an entire batch.
Ditto. I’m so glad you’re going to opening a new window. Often, I just reopen your site from scratch because I’ve gone off on a trail.
The video was impressive, but I can’t say I’m completely convinced. He was doing a no-brainer job. For me it’s faster to batch sometimes. For example, on processes I have to think about, if I batch, then by the second one I don’t have to rethink the process, and by the third one not much thinking. Also, when I have to stop and leave my machine in a process (like to iron or cut or change from serger to sewing machine) it’s faster if I’ve got several items going at once. On the other hand I do get a fabric pile up sometimes, but I solve that by tossing half finished garments in a pile on the ironing board!
If he had to get up after each envelope was ready and drop it in a mail box (the equivalent of changing stations from sewing machine to ironing board, for example) then I wonder which process would be faster.
Also, I sure wish he’d put the envelopes over by his left hand with the flap pointing towards the papers. That would have sped up both processes.
Okay, I’ve over-thought this enough for tonight.
Have a great weekend everyone.
Marguerite
Mike, RE: catching defects, batching etc; I made all those points and any caveats through linking in the entry rather than repeating myself ad nauseum. I realize this is your first comment to this site but maybe you’d find some common ground regarding your points, via the examples left via links in the entry.
Yes, yes, new windows are much better! I often am torn between following your always-interesting links and not always being able to come back easily to where I was on your site. I vote a big YES.
regarding new windows: already in the habit of using right click/open in new tab so as not to lose the original page :)
that video was cool and unexpected. always thought batch was faster! very good to know!
I prefer the links opening in a new window. It saves the step of right-clicking and choosing that option, which I sometimes forget to do. I think it will be an advantage to leave the F-I page up for an easy return to it.
Great post, thank you.
I was thinking the same thing as Marguerite though. Maybe not being as efficient changing machines, getting up, walking to the next station and all. But definitely worth it quality control-wise.
There are definitely times where batching is necessary. This comes back around to Kathleen’s example of pulling a serving of corn. Some things, inherently, require time to manifest and its gestation simply cannot be reduced any further. That’s OK. The processes just need to be designed/orchestrated to work with it.
Now, coming back to batching in the work environment. If you’re a single operator workshop, batching rears its head more often. Case in point: sub-assemblies. If you’re making jackets, you’d probably batch all sub-assemblies (e.g. sewing, turning and pressing the breast welt, creasing welts, lining, turning and pressing pocket flaps, attaching bridle tape to chest pieces, etc.) before assembling them into the coat fronts. The point is to get as close to single unit flow as possible.
RE: formatted links. I’m a Firefox user. So opening multiple tabs is ingrained. But, judging from the rest of the feedback, the formatting looks to be a hit!
Another ‘yea’ vote for links opening in a new window. Good kaizen.
I will have to try this batch/lean experiment on different processes. However, I can’t see washing clothes one by one as more efficient than doing a batch.
Regarding changing work stations: First, I’m going to point you to observations of single-piece flow in real apparel manufacturing:
SPESA Trip Report Eric
SPESA Trip Report Kathleen
How to Get People to Change
Batch UPS and Modular Batch
Fit Couture Lean Manufacturing Implementation
Back From Houston
Clothing Manufacturer Fit Couture Trip Reports
Clothing Manufacturer Fit Couture Trip Reports 2
It works!
Second, there is no “get up and move to the next machine” because the work is done standing up. There is some expense to the initial setup, but you get results right away.
Third, in the linked video, notice how quickly the first customer gets their order. In batching, the first customer doesn’t get their order until nearly all they way through. He starts at 1:31, doesn’t stamp the first piece until 4:49, a total wait of 3:18 if he got up to mail it then. Of course, he doesn’t because this is batch! So the first one won’t be mailed until after he completes the last one, a total wait of about 3:47. And the 2nd, third, and nth customers get their order in about the same time.
In one-piece flow, on the other hand, the first customer gets their order in 0:14. Fourteen seconds. The second customer gets their order about fourteen seconds later. And so on.
Now, imagine if you had a special order or some other unplanned event come in right in the middle of either of those processes. In batch, the entire stack would get put to the side while the special order is processed, but in OPF, the first few go out the door, you process the special order, and then you return to the original sequence. Only a few customers are effected.
My experience is that “special orders or other unplanned events” happen nearly every day. Even the “special orders” get interrupted by subsequent “special orders”. The whole shop becomes one big pile of work in progress (WIP), nobody is happy, morale suffers, and you spend your whole day putting out fires and making excuses to customers. One-piece flow allows you the flexibility to handle those “special” events (which aren’t really that special if they happen that frequently).
Regarding those “unavoidable” batches: I will allow that things like washing may be more effectively done that way, and of course cutting and some sub-assemblies. Even Toyota makes sub-assemblies on lines that run parallel to the main production line (though the are synchronized precisely). But I would be careful about making the excuse “but I’m only one person”. After all, each person in the lean implementation is frequently doing exactly what the one-man shop would be doing.
I think we get comfortable doing things in batch operation because we fool ourselves into thinking that we are increasing efficiency with each repetition. When you force yourself to confront the problem — “How could I do this effectively in a one-piece flow?” — you will frequently find that you can do the whole thing much faster. That’s the point! You discover the little operations that slow things down and either streamline them or discover how to eliminate them altogether. In batching, you learn to speed up each operation, but in OPF you learn to speed up the whole. One is local optimization, the other is global optimization.
Marguerite said a lot of what I was thinking and have experienced. I will add that my mother, perhaps a natural efficiency expert, always said that if you have to handle something more than once you are wasting time. So the envelope gets handled, in batching separately three different times and the operator has to check to make sure that the orientation of the envelope is correct for the stamp.
For times when batching works–such as sewing all the collars then pressing the seams–I can’t imagine that sewing each collar and getting up to press or walking over to press each would would be faster or more efficient. Perhaps if you turn and the press is right there next to you. But in some cases you can sew one item and then next and the next without stopping. So batching does makes sense sometimes.
I love the links opening in a new window! I will go off on a trail and forget why I started down the trail in the first place.
Eric makes good points, but I still think if you’re a one-person operation, it’s hard not to batch some things. Even walking to the next machine isn’t efficient for one person. But if you have several people, then I can see the flow of each person moving along as at Fit Couture and finishing the whole garment works, but then you have the next garment right behind.
I have garments I can completely finish at one machine with no interruptions, and when I have a pile of those, I definitely do them one at a time. It’s faster than handling each one a bunch of times & keeping them separated. But then I throw them (literally) over onto the ironing board, and I iron all at once (batch ironing).
I don’t have the customer problem as I’m not doing 1 and handing it off to a customer.
Anyway, we all can learn from these ideas and try to be more efficient.
Marguerite
I don’t find the video believable – it makes me doubt his “kaizen” approach, not trust in it. He set up his work station to have the worst workflow possible when he was doing the batch job. Then, when he switched to the lean approach, he used a less inefficient workflow (I can’t go so far as to say efficient, since it was still absolutely far from it). How is this comparable? Moreover, he is the slowest paper folder on this earth. I think his ideas are credible and applicable in many instances, but this video gave me less confidence in his methodology, not more.
Ouch! By turns I’m shocked, dismayed and not a little embarrassed. Obviously, I need to do a refresher on lean for people who’ve joined us recently or tuned it out before. I’m really surprised people think this is a “theory” to be debated, or it being a matter of opinion. I am really surprised.
One doesn’t *need* to trust it to believe in it. This isn’t faith based manufacturing. Do the same experiment yourself and time it. If you’re having a hard time “believing” him, I’ve got a whole laundry list of stuff (that works) that you won’t believe either. “His” kaizen ideas arent’ anything new. Might want to peruse the Lean manufacturing category over there in the right sidebar for three years worth of it (not that it was my idea either). Like I said, I obviously need to do a lean refresher…I am surprised and not a little dismayed.
And as I said in the entry, we’re hard wired to prefer batching. I guess shouldn’t be surprised to see people continuing to defend it now. Easy doesn’t equal efficient. If you must justify your batching, defend it as being easier for you but it’s not more efficient.
I actually agree with Jennifer. Not that I think that Lean is something new or debatable, but that if Ron Pereira (or anyone else) wants to make a video to visually demonstrate that one process is more efficient than another, he shouldn’t compare really terrible batch processing with okay one-flow.
Even if he’s right, he looks like he’s cooking the books – especially if he then comes up with a number to make the whole thing look scientific: we can all see that the number was rigged. (He gave us a 1:01 time saving, 3:47 vs 2:46. I would have been completely convinced of one-flow *and* Mr. Pereira’s credibility as a one-flow teacher with numbers of, say, 0:46 time saving, 3:22 vs 2:36.) If I were interested in pursuing lean processing, this video would stimulate me to want to learn more… and also ensure I looked to anywhere but Mr. Pereira for further information.
I love a new and separate window to open when I click links!
The one-piece flow demonstration inspired me to time and test some of my own processes.
As far as the links are concerned: I don’t care. I’m always using right click to open in a new tab, so for me it’s no difference.
For the batch vs. one-piece flow I’m not sure what the intention of the video really is. As some others mentionned before, it is a nice illustration of an idea but there are so many small differences in the layout of the two different processes (it also seems to me that the white paper was thinner, therefore quicker to fold than the red one), that I cannot call it convincing. Unless I do believe in the idea of one piece flow being better than batching already…
As a counter example from a completely different field: For some years I was making all my christmas cookies together with a friend. And in one day we were constantly producing three times as much cookies (in number and in varieties) than each of us produces alone. Just because sometimes ist saves time if one person prepares the dough for the next load while the other is getting the first load from the oven.
So I think having more people doing the job could change the result of the simulation in the video. Or maybe not… but the video doesn’t show that. It’s just not a “valid” simulation for any process done by more than one person, but a simple illustration of a theorie.
(But… I also think that 100% batch processing is not the best answer to all production needs. If the video is meant to make people just think over their production process it has definitely good intentions. Because, as often in life, there is not one standard solution to all needs, but everyone and every company has to find the one that fits their needs best.)
(And batching is more boring than one piece flow…. that already makes me avoiding it. But I’m not a professional garment producer.)
I was not saying lean manufacturing is a bad idea or something whose credibility I doubt. I simply doubted the video. You can have the best process in the world, but if the way you present it doesn’t get that point across, then people aren’t going to buy into it. I buy into lean manufacturing because I’ve read a lot about it, not because of this video. If this video were all I’d seen, I’d think the whole thing was a crock.
My background is epidemiology, which is all about backing up medical theories with facts. Same thing applies – your theory can be dead on, the best approach in the world, but if you set up your experiment to give bad results or useless information, no one is going to believe you.
Interesting.
As I watched the video, I didn’t see any particular bias in the way he conducted his operations. I assume that both the batch and single piece flow could be further improved, but that wasn’t the point of the video. I’ll have to take a look and see whether I can detect any if I’m specifically looking to find it.
Someone mentioned that they could see single piece flow working in an environment with multiple workers (like here at Fit Couture) whereas it might not work as well in a single employee example.
I would counter that the opposite is true. At least in our case (and logically, it should hold true in most other cases as well), we would achieve our highest theoretical output by maintaining a 1:1 ratio between physical manufacturing lines and employees. We would probably still treat them as a team for bonus purposes to maintain the shared goals and camaraderie that TSS fosters though.
The “bumping” that occurs during single piece flow is wasteful. Its a big improvement over batch mode, but there is still waste there.
Any time our line is short handed, efficiency is higher than if its fully staffed. Indeed, I’m starting to suspect that our workers have figured this out and are starting to game the bonus plan a bit by minimizing the number of hours per day that the line is fully staffed (i.e. one worker comes in early and one stays late) to take advantage of that. In as much as it doesn’t force me to work longer days than I’m already working, I don’t really care. As the team efficiency increases, the stitchers make more money while our actual cost per unit drops.
Okay, I agree, evidence is compelling. How do we account for the ~60 second difference in the two processes? (Which I contend is the wrong comparison, more below).
Average time to fold:
Batch: 9 s
Lean: 8 s
Average time to stuff:
Batch: 4 s
Lean: 3 s
Average time to seal:
Batch: 2 s
Lean: 1 s
Average time to stuff:
Batch: 1 s
Lean: 1 s
So, over 10 repetitions, he got a total of 10+10+10=30 seconds of advantage from the shorter time to fold, stuff, and seal.
The shorter fold time could be due to thinner paper, I’ll grant that. Or to the fact that he seems to get better as he goes (he starts out with times of 9-10 s, but ends with times of 7-8 s). The shorter stuff and seal times, though, are due to the fact that he is already holding the item from the previous step. He gains 1 second each time from not having to find and pick it up. That’s part of the point, and another lean production trick (SMED, the one that made Shigeo Shingo famous), so I contend that it’s unfair to count those against him as if they were a parlor trick of some sort.
Still need to account for 30 seconds, though.
He loses between 2 and 5 seconds every time he moves the pile around between steps. Also, he has to manage the pile several times during a task, something he doesn’t have to do nearly as much with OPF. This also has a factory corollary: storing, moving, retrieving, and looking for WIP.
But those are the wrong numbers to compare. The real advantage, though, is the fact that he is knocking out a complete product roughly every 15 seconds with OPF. Every 15 seconds, the lean manufacturer fills another order. Every 15 seconds, he has the opportunity to inspect WIP and final product for defects.
Heck, let’s even spot the batch production method the 3 second difference (most of which is legitimate gain) so that they both average 18 seconds. The lean producer would be still be fulfilling another order every 18 seconds. The Batch producer doesn’t get any orders filled until 3:47. What if they were hours rather than seconds? With 40 hours in a week, that means that the lean producer is shipping twice a week while the batcher is shipping every 6 weeks. Do you like the idea of cash flowing in twice a week, or every 6 weeks?
Eric, you’re assuming you’re shipping an item as soon as it’s ready. I’m usually working on several orders at once that have the same ship date.
I think we’re beating a dead horse here. Given ideal conditions, lean is faster, but we don’t all operate in ideal conditions, and we have to combine batch and lean.
I don’t have enough machines to have one set up for everything I do. Sometimes I need 1 needle in my serger, sometimes 2. I’m not going to stop in the middle of a skirt (for example) and change the needle set up (which always requires testing the stitch, an extra step every time I change the threading) if I’m working on 3 skirts. I’m going to do all the 1 needle stitching on all of them, then change to 2 needles and finish all of them–except for the last step which requires me to go back to 1 needle. And there’s the sewing machine–same problem, different thread for sewing in labels than for the finish work on some garments. So I do a bunch of labels all at once. And then there’s buttonholes–same sewing machine, different set up.
It’s a pain with all the fabric handling…but until I can have a bigger workspace with all the machines set up as I need them, I just plug along being as efficient as I can be.
Bottom line for me is that I like lean when I can do it, but it’s just not always practical. Sometimes it’s just a matter of which I hate least–handling batch fabric or changing set-ups to do one garment all at once.
Kathleen, don’t be embarrassed. I think we all understand lean, we strive to be as fast and efficient as we can, and the interest in Mike’s setup is evidence that if we could have the ideal shop, we’d have lean! I know I would.
Marguerite
Thanks for the passionate dialogue everyone. I love this type of discussion as it helps me learn.
I may do another video where we make PB&J sandwiches in batch and then one piece flow using more of an assembly line. I may get my kids in on the action. We can then donate the sandwiches to our local soup kitchen teaching the lean principle of “respecting people.”
I am curious if anyone has ideas for how to compare one piece flow top batch production using aspects of your business (e.g. fashion, sewing, etc.). If you have any ideas please do share.
Thanks again everyone. All the best.
I don’t dispute the lean manufacturing value, but I think the example chosen is poor because it presents the fallacy of the false dilemma: his two processes are not the only choices.
As a long-time secretary I’ve stuffed my share of envelopes, and neither of those choices represents the way I was taught to do it, in a few particulars.
The “downside” of lean is that it often requires a significant investment up front.
The lean solution to your machine issues would be to add more machines (or switch to machines that could
multi-purpose without any delays).
Its a straightforward calculation to determine whether the efficiency gained would be justify the cost. (Though, in our case, America’s 21st had to both calculate the equipment cost upgrade and expected efficiency improvements.)
Now I’m curious. What is the better way to stuff the envelope?
Mike C: With single-sheet flyers you can grab 5 or 10 and fold them with a quick crease, then separate the leaves as you grab them for insertion. Stacking the envelopes flap-up, oriented so the opening faces the hand that grabs the flyer.
Then, too, our postage meter has a water reservoir and seals as it applies postage.
A lot of interesting comments. The one thing nobody has really mentioned is that the idea with Lean is to work “towards” one piece flow with one-piece being the optimum. There are many reasons a process might not be able to work at one piece flow; distance between processes being a biggie. The idea with Lean is to remove the “wastes” (distance being one), so that you can get closer to one-piece flow. In the example where you wouldn’t want to have to walk over to the iron to iron one piece, the answer would be to somehow connect the processes so that distance would not be an issue. The “Big Picture” of the video was to show how one-piece flow would reduce lead time by removing motion wastes that added to it. If you have to pick up a piece more than once, you are wasting motion. In his example, he keeps the part in his hand at all times through the process; thus reducing the motion and transportation waste associated with batch. Remember though, you cannot always go directly to one-piece flow. There are lots of rocks under the water (inventory levels) and if you lower the levels too quickly, you are sure to hit one of them. Remove the rocks first, and then slowly reduce your batches.
The difference in time comes form how often a piece is picked up and then laid down again, not form how fast the actual work is performed.
As little experience as I have with production, my rule of thumb is – if i can keep going without having to lay the piece down, then keep going, if I have to lay it down anyways, i go for repetition of the process with a new piece.
eg: if i have to go and press a seam before being able to continue sewing, i move on to the next piece and finish the batch before starting a new process-cause it involves walking a few steps. Mind you-the walking would keep me more fit :D