How to sew faster pt.3

This entry could be summarized by saying maybe industry doesn’t sew so fast after all. Pam gives me the opening I was looking for with her comment from Tuesday:

The black shirt most recently pictured on my site took me about an hour to actually sew, after cutting and fusing. Yes, the seams are flat-felled. Most home-sewers/enthusiasts think that is amazingly fast. Professionals may think it rather slow.

My response:

This is an interesting point Pam. I plan to write about the real costs of handling, a concept that is not measured or included in total production time in the typical factory.

In other words, your sewing “time” may not be as inefficient as you may suspect as compared to a typical shirt manufacturer because they are not using the same ruler you are to measure total time spent. The failure to include the time and cost of handling is the summary argument for employing lean manufacturing of which the tenet (in apparel) is one person sewing one item start to finish -like you do.

Industry and enthusiasts use two different rulers to quantify the length of time it takes to sew something. Our definition of time spent is entirely different and more tightly controlled. In home sewing, time is (comparatively) ambiguous -that’s not a criticism. Home sewers count the total project time, perhaps starting with cutting and stopping the clock after the final pressing is done. We don’t do that. Or we mostly don’t.

If you ask someone in the industry how long it takes to sew something like a sport coat, they’ll tell you something on the order of 28 minutes and you will ooh and ahh and be very impressed. We like saying this, puffing our chests out a bit. Yes we do. Problem is, it isn’t really true, not that we’re lying, we’re using a different ruler. The below specifically refers to a batch sewing process (meaning one person does the same operation on a bundle of cut goods over and over) and what many (enthusiasts included) think is most efficient. It isn’t but that’s another story and the meandering end point to all of this.

When we say a sport coat takes 28 minutes to sew, it means we are paying for 28 minutes of a sewing machine operator’s time to sew that coat. That doesn’t mean they’re doing work they aren’t paid for either. It means we’ve done the time and cost calculation of every single step of the sewing process -just the sewing part. We call this SAM or Standard Allowable Minutes. For the sake of simplicity, let’s say the average sewing operation takes 20 seconds (that is actually quite long) or three operations per minute. That means there are nearly 90 different seams sewn on a sport coat (there are actually more).

Now, just because we’re paying for 28 minutes of time, doesn’t mean it takes that long. You cannot throw a sport coat into a sewing line and come back half an hour later to start collecting jackets feeding off the end of the line. The very idea is laughable. The time spent handling the goods, schlepping it from place to place is either invisible or not quantified in our definition of “sewing time” because the cost of handling is perceived to be of a cost so low, it can’t be quantified. There are two kinds of handling for the purpose of this discussion.

One is in the sewing of pieces themselves and the time and cost of handling are rolled into the piece rate of sewing operators. Pick up a piece, pair it, sew it, fold it and lay it aside. It must be folded and lain neatly to reduce pressing time later on. Done with the bundle, you tie it up again, attach the bundle control tickets and place the bundle on a side table or trolley. This is where the second kind of handling comes in. A “floor girl” (or conveyor system) will pick up the bundle and transport it to the next person in the sequence because work operations are not necessarily adjacent to each other. Floor girls walk all day long, and tend to be very thin. Seriously. Anyway, the second kind of handling cost is reflected in the hourly wage paid to a floor girl. As it is low skilled, the cost is so minimal it can’t be calculated into the cost of product except as administrative overhead.

Another way time is wasted and not rolled into the total time calculated as home sewers would think of it, is waiting or resting time. There’s bottlenecks where work in process (WIP) is just sitting there with no one doing anything to it. In factories, this is perceived as mostly “free” but it’s not really. There’s costs associated with it in storage, keeping it clean and neat, inventorying the bundles to feed through the line, or rack space to hold jackets waiting for pressing, to say nothing of space requirements and utilities. It would be very interesting to tag one given unit of a product with a timer to see how long it took to process from cutting until it was completely finished. I’ve heard that some plants making commodity items take three weeks to process something start to finish. However, a three week processing time would not keep a plant manager from boasting it took them seven minutes to make the item, something a home sewer could finish in an afternoon.

So, what is “sewing time”? In the opinion of lean manufacturing adherents, sewing time isn’t a good measure of costs, the total time spent in process including handling and waiting around should be included. In this light, industrial sewing isn’t nearly as fast as the previously quoted 28 minutes. Generally, I would guesstimate that products that took longer to sew (28 minutes) have less wasted resting time than commodity products. If it takes longer to sew, it costs more (materials too), there are fewer of them so more is invested in the relatively fewer pieces and you need a faster return for your efforts. These factories are usually smaller too. This is why the opposite is also true too; commodity plants are really huge, they need the storage of WIP. Relative “efficiency” is subsidized by huge plants in low rent areas here or offshore.

In summary, I think an ambitious target for enthusiasts to shoot for is the total time it takes a good sample maker to make a product start to finish with a single needle machine and overlock. A good sample maker can make the 28 minute jacket in under three hours. Again, this doesn’t include pattern or cutting time, this is done for them. Also, while the sample maker does all of the interim pressing, the finishing (button holes, buttons, final pressing etc) is done by those work stations.

[Not wanting to bore anyone with a repeat on the evils of batch processing, aka the waste associated with “handling time”, see this entry for an impressive experiment you can do at home and then this one.]

Related:
How to sew faster pt.1
How to sew faster pt.2
How to sew faster pt.2b
How to sew faster pt.3
How to sew faster pt.4

There are 10 comments. Leave a comment

Leave a comment

Your email address will not be published. Required fields are marked *