Lean Manufacturing

A discussion of how to eliminate wasted time and effort on the shop floor. October 19, 2006

I've been doing a little research on lean manufacturing and would like to start implementing some of the techniques in our shop. One thing that bothers me is batch sizes. We use Cabinet Vision to create and optimize our cultists, and then cut everything out on a CNC. Doing this eliminates a lot of waste material per job. If we go to smaller batches to increase work flow, the amount of waste per job goes up significantly. For example, we are about to start a job that looks like will take around 35 sheets of 1/2" material if we batch the entire job at once. If I break it down to smaller batches, like 6 to 8 cabinets at a time, we will end up using around 42 sheets of material. This seems to be making more waste for the sake of increased work flow. I was under the impression that a major part of lean manufacturing was eliminating waste. I would really appreciate any insight here.

Forum Responses
(Business Forum)
From contributor A:
Will CabinetVision allow you to nest your parts to partial sheets? If so, save your cutoffs and nest new parts to the partial sheets. This may work for you.

From contributor B:
I'll take a shot at a lean manufacturing definition. This is from my understanding of the process so there is no guarantee of accuracy. The goal is to get the most product out the door with maximum added value to the customer with the least amount of effort to produce it, in as short amount of time as possible.

Efficiency is the key. If you can draw, detail, engineer and nest all the cabinets for a job at once without the shop waiting around for information then that is obviously very efficient. But around here we very often are bogged down by a single cabinet or group of cabinets that could potentially hold up the whole job. Also it can be hard to produce all of the shop information for a large job in a timely manner. So it is often more efficient in the long run to release segments of a job to production. The cost of waste material can be very cheap compared to unproductive shop time, or worse yet, re-dos. I would say that the optimal batch size is the largest batch you can handle while maintaining control of quality, speed and standards. I think that work flow has a huge priority over waste factor. When have a good work flow then you can focus more on using material more effectively.

From contributor C:
In cabinetmaking, most of us have a custom or custom stock hybrid type of manufacturing process. Lean scenarios for this type of manufacturing are a bit of a different animal than they are for pre-engineered product scheduled in batched production runs. The optimal batch for production efficiency can be elusive. It may require you to filter the batch to contain a particular number of units made up of certain types of product to find an efficient batch for both material yield and production. If you have the current version of CV Solid it should have the controls available in the NC Center to let you get pretty close.

It's true that material waste is usually less expensive than lost productivity but often times, depending on the footprint of your shop, dealing with more waste material and trying to recover offcuts for production can throw some of the lost efficiency back on the labor side. When trying to apply lean strategies to a small to medium sized custom manufacturing business it can become a one step forward, two steps back type of game if all things are not considered. The best measurement I have found for results of lean manufacturing in wood products seems to be throughput. I was taught that throughput is the measurement of all resources expended between the time you begin to commit resources to a job and the time the job is complete and paid for in full. The measurement of expended resources should include materials, time, labor, cost of inventory and overhead. Tracking all of that can be a big job. In many cases it requires setting up systems to provide metrics on specific aspects of the business on a per job basis that were not previously required. Sometimes lean systems end up providing less than expected in real R.O.I. because of incompatibility with aspects of custom manufacturing and the burden of new processes required to monitor results. That's one reason why business process management software is beginning to trickle into the wood industry more and more these days. It provides a skeletal system for monitoring and managing a business that can be customized to fit the way a business operates and identify opportunities for improvement based on that rather than applying a blanket strategy like lean to the business. I hope this opinion helps you in your efforts.

From contributor D:
I don't see a problem with processing the panels in one whack, and then fabricating in batches from that point, assuming that you have the space in the shop to accommodate this, and your employees have good discipline about not starting a new batch until it is time to do so.

There are costs to having parts sitting around waiting to be used, as I'm sure you're aware; it's money tied up in inventory that reduces cash flow; there is the opportunity for the parts to be damaged before they get used, etc. I think the important thing is to realize how much space you are taking up with those parts, and how much money you have tied up in them, and what the potential pitfalls are of having them. Then you make a value judgment based on that. As they say, there's more than one way to skin a cat.

From contributor E:
Those 35 sheets take 46.66 sq. ft. of space. If you were talking about 35 units of 50 sheets each you should consider lean manufacturing. At least for the cutting aspect this is making a mountain out of a molehill. Cut it all at once and keep all parts neatly organized for assembly in batches.

From the original questioner:
The more I think about what's happening in our shop and from responses I've gotten here, it seems that the real problem is scheduling, not batch sizes. It's not uncommon for everyone in the shop to be pulled off other things in order to get one job done. Once that job is done, everyone is standing around waiting for the next job to be cut.

From contributor F:
The real problem is batch size. Building something before it is needed is the root cause all of the other problems. The optimum batch size is probably best correlated to how big the package is when you carry it off the bench, onto the truck and into the house. Your goal is to become the low-cost producer of that package then become the low cost producer of the next package permutation.

Lowering your batch size will eventually lower how much work you have in process at any time. As soon as you lower this work in process (WIP) it is going to become real obvious where your real problems are. It's probably not the extra 13 sheets of material. Piles of WIP are hard to see over or through, and as such they tend to mask the real bottlenecks. If you have a melt-down at one work station and you have extra parts to pull from, it's usually easier to grab another part than figure out and fix what really caused the problem.

Lowering the amount of WIP and moving from a push system to a pull system is something you will want to do gradually. Eventually you will come to realize that Mrs. Smith is just the client for your project, and the real customer for your current process is the next process in line. If you can only hang one door at a time, building 30 doors just means you will need a place to store 29 doors. You might want to consider contributor D and Es advice. If bigger batch sizes work best at cutout, develop a storage and retrieval system for the cut-out parts. Developing the kind of discipline it takes to not launch into unnecessary WIP could be just as simple as releasing the construction documents one cabinet at a time.

From contributor H:
My view of lean is that it is an attempt to eliminate waste and maximize profits. If the material cost goes up by some shifting of quantities more than that much cost must be gained somewhere in the entire system in order to justify the shift. We struggle with the same problem. Much of it is hard to measure exactly without spending too much time collecting data and analyzing. We have attempted to use day production quantities as measured by what ever is the slowest link in the process. I hate to say it but the office is often the bottleneck. The major machine areas have more capacity than can be used in much of the rest of the shop but you can't buy a .788 CNC router. Our panel processing area is all on roller conveyors. The transfer cars allow parts to be shuffled as required but the lineal ft of conveyor sets a limit on how much WIP can be sitting there.

We limit how high the piles can be and don't allow stacking parts on to carts or pallets. At the last part making station the parts are collated into cabinet parts in the reverse order the parts will be made into cases. Each assembly station has two 10' sections of conveyor leading to the bench, which makes it easy for the shop foreman to see what progress is being made at each bench. We have purposely limited the amount of space available for WIP cases after assembly and before laminating, finishing or final assembly. That provides another automatic stop, forcing a redistribution of labor. It's not a perfect scheme but it has helped a lot. I try to keep from starting any job until as close to the ship date as possible, that way the last minute changes by the customer can get made without work redo. I still have problems with that concept because of schedule changes, vendor and employee problems. I wish general contractors would let you know they are behind before the truck is being loaded! We photo and measure the site as late in the process as possible so that as much of the interior is complete as possible before we start production. Most shop drawings are already complete by then so they can be released to the shop on time if there are no major onsite problems. Before we had good software this was a risky deal but now with parametric software changes go very quickly. The entire business is a lot like solving a puzzle with a moving answer. Anyone interested in Lean should read "The Goal", then books by Dr. Demming and Tachi Ohno.

From contributor G:
To contributor H: You seem to be pleased with your current software that allows for the parametric changes you mentioned. I'd be curious to know what you are using and what your primary product is.

From contributor H:
We use Microvellum. We also have an ERP program that I'm hoping to replace as MV gets more of that integrated into their program. Our products are commercial interiors, store fixtures, medical facilities, moldings (straight and curved).

The comments below were added after this Forum discussion was archived as a Knowledge Base article (add your comment).

Comment from contributor J:
Lean manufacturing is a sound principal when properly implemented. Unfortunately, the focus is often misplaced. Because batch size and WIP are easily measured, these metrics frequently become the goals. The true goal is to increase throughput, which is most easily measured in shipping dollars. To do that, you need to look first at work flow.

I would suggest doing some reading on theory of constraints. Find your biggest bottleneck, and subordinate everything else to using that bottleneck at maximum efficiency. That will lead you to the proper batch sizes and proper amount of WIP.