This is the third in a four-part series on lean manufacturing for custom shops. Part 1 covered the importance of a clean and organized shop; Part 2 focused on identifying common tasks. Part 4 will explain how to get employees to “buy in” to lean manufacturing techniques.
As stated earlier in this series, material handling is one of the main areas where custom shops can find waste. And examining your “batch size” can be very illuminating.
From our days of using table saws, molders and such, we have learned to set up a tool and cut every part needed before breaking down the setup and cutting the next set. For most of the cabinet shops I have worked in, or with, that same mindset continues even on the CNC that they use. But have you thought about why? I advise shops to mill all the parts for each cabinet at the same time. For many, this sounds like heresy. But let’s look at how it really works.
You work through your kitchen project in whatever software system you are currently running and, when everything is complete, you output to the CNC. Your operator then loads sheets, runs the programs and stacks up the completed parts ready for the next station. Those parts are typically in no specific order; the computer was striving for maximum material yield with the parts in play. These disorganized stacks of parts have to be moved to the next step in your operation.
The edgebander is usually the next stop, at least for many of the parts, but they are mixed together with the parts that will not be edged, so they need to be sorted. Unless you have a lot of floor space and very good carts or racking, these parts will be stacked in piles and, coming off the edgebander, will go back into piles. And every time they are sorted, time is spent identifying which parts are which. Now the piles are sorted again to get the parts for each cabinet together for assembly.
Most parts, by the time they reach assembly, have been handled several times and are quite possibly damaged. And unless you have a really good system for marking the parts, a lot of time was spent just identifying them.
Dealing with mistakes
And what happens if a mistake was made in the original programming? Let’s say that, for some odd reason, the shelf holes do not line up left to right. You know stupid mistakes happen more often than we care to admit. By the time this is discovered in assembly, most, if not the entire, job has been cut and you have stacks of parts that are unusable or need to be reworked.
So let’s look at that same job run as a single-unit batch. Virtually every software system can be set to nest by cabinet (unit) as opposed to nesting for maximum yield. Your CNC operator now runs the programs and the parts for Cabinet No. 1 can be stacked together as a set and sent immediately to the edgebander, while the CNC continues to cut parts. Now the edgebander is running at the same time as the CNC, just a step or two behind, and the first set of parts moves on to assembly. When the problem is detected with the line borer, the operation can be stopped and fixed. The CNC was only just cutting the parts for Cabinet No. 4, so the remainder can now be cut correctly. The number of unusable parts needing rework is much smaller. You just saved a bunch of time and money.
And if the programs were perfect to begin with, you still save time and money sorting and stacking parts. No real need to fuss with extensive marking of parts since each batch has two sides, a deck, toe kick and stretchers. No more large, heavy and unwieldy stacks to be moved around and the parts are less likely to be damaged without all the handling.
I have a client that supplies counters and sales stations to large chains. Every job is different, but it might entail as many as 400 units. If they cut all 800 sides for the big job first, there are huge stacks of sides to be stored while the other parts get cut. Single-unit batching was the solution. Because each set of parts moves through the shop as a unit, they can break into their production as needed to get a smaller job through without interrupting the larger job by much at all. The only place that anything tends to stack up is in shipping where storage does not interfere with production.
Is single-unit batching always the answer? Of course not. But studying your particular work flow can help you identify where batch sizes can be reduced to a lowest common number. For example, if you produce face-frame cabinets with a single frame that spans two or more cabinets, then it probably makes sense to cut your parts in sets by frame so that they can be built together.
Large batch sizes used to be determined by the amount of time required to set up a machine. When everything was manual, it made sense to cut as many as possible to spread the cost out. But your CNC machine does not really need to be set up in the same way. And the digital revolution is moving deeper into your shop. Modern edgebanders are computerized and can switch from one type of banding to another in seconds. Even table saws are getting digital rip fences and sliders can now be equipped with powered, digital rip and crosscut fences that are as accurate and repeatable as a CNC.
In your shop, what are the common tasks jigs and fixtures can be used to help reduce batch size? Years ago I built a multistation router table for making paneled newels. Once the setup was right, it was never changed except for sharpening. Any time I needed a newel I could cut it without waiting. For another client, I recommended a dedicated router table setup with a drawer lock bit installed for cabinet box production. Since the length of the stock determines the box size, the drawer lock setup never changes. One drawer or a dozen, each drawer takes the same time. For oddly shaped parts that need to be made regularly, a template was kept handy to rough and trim them to size quickly.
I know many of you reading this are still skeptical. But I challenge you to quietly follow a couple of jobs through your shop. Look for stacks of parts. See how many times each part is picked up, moved and put back down. Look for ways to group your production to maximize flow and minimize handling and you might be surprised at how much time and money you save.
Ralph Bagnall is a woodworking consultant and can be reached through his website at www.consultingwoodworker.com.
This article originally appeared in the December 2011 issue.