Two Workers on One CNC?

Is two men on a CNC machine one man too many? Here's a long, informative discussion on improving labor productivity with CNC equipment. October 25, 2007

About 1 1/2 years ago we purchased a CR Onsrud CNC router. It has a 5' x 12' bed, 1 router and a 12 place tool changer. We currently have two guys running the machine. Most of the programming is done by the front office guys who run Cabinet Vision and output the parts. So all they have to do in the back is double check the g-code to make sure the z level doesn't go below -.005.

The one CNC guy, who I will call Bob, is responsible for programming all of the single parts and/or redoes that need to be cut. He uses Alphacam to do all of the drawing. This, of course, does take some of his time.

The other guy, who I will call Bill, also has some other responsibilities. We use hardwood shelves for the wall cabinets and he has to cut them down, among other things. Together they also stack the cart. When a Cabinet Vision job is completely cut, they load up the cart with the entire job, or as much as will fit. They start off with the highest number and stack them accordingly. With the entire job being optimized, it takes a bit of sorting to bring all the cabinet parts together per number. During this time, the CNC is usually not running.

It all comes down to unloading the parts and then reloading. With two guys, you can take all of your parts off, blow the spoil board on, and get the next sheet running a lot quicker then you can with one person. But then again, while the sheet is running, you have two guys waiting. (Maybe they are doing something else in between, but you can never be as productive when you are always being interrupted.)

What have all you out there in WOODWEB world discovered? Should we insist on only one man at the CNC, or is there benefit in two?

Forum Responses
(CNC Forum)
From contributor D:
If you don't know what they are doing while the machine is running, you need to pay more attention to what is going on in your shop. One person should be able to load and unload this machine, then the parts can be sorted somewhere else. The operator may need some assistance every so often, but two people at the machine all the time is excessive. Loading heavy sheets can be a chore for one person, but unloading is smaller pieces; the bottleneck seems to be in the sorting. Why is Bob doing so many redoes at the machine? How many people are up front working on faulty programs? The only place I see more than one person running a machine is at trade shows, where it seems like it takes 5 technicians to set up something that the salesman tells you will take no time and make you tons of money.

From the original questioner:
There aren't that many redoes; it's more that he is programming single parts like valances, fluted boards, etc. I agree the bottleneck is sorting parts. But to load them all on a cart and then unload them again for sorting seems to be additional handling. As it is, they take them off of the table and set them up against the wall till the whole job is cut out. Then they start from the highest number cabinet and put the parts flat on the cart. We are looking at the possibility of building a cart with vertical slots that would allow us to put the parts of each cabinet in vertically so they could be sorted as we go along.

From contributor C:
Yes, but from your post, I see other areas I would address as well.

1."Most of the programming is done by the front office guys who run Cabinet Vision and output the parts. So all they have to do in the back is double check the g-code to make sure the z level doesn't go below -.005."

Contact CabinetVision Techs and have a "Z Filter" written into your post processor. They did this for free for us and since then, CV can't hit the table.

2. "We use hardwood shelves for the wall cabinets and he has to cut them down, among other things."

A simple UCS needs to be written and CV will yield and show a shelf edge to the shelf, cut all those shelves to the correct size for you, and even rabbet an edge, if you like, for that hardwood.

3. Programming at the machine on the fly may seem like a cool idea, but it also opens the door for your cabinetmakers/assemblers to interrupt your operator and slow him/her down. In our situation, I am the programmer, I use CV and AlphaCam and prepare both CNC packages and shop packages for every job.

My primary operator's sole job is to load the programs from the server into the CNC's controller, pull the materials, load the sheets onto the bed, set the vacuum, and hit the green button. When the program is done running, he labels all the parts and sorts them into two groups - those that need something done to them and those that don't. Examples of the first type are parts that get banded or secondary operations.

After he has all the parts quickly out of his way, he cleans the table and loads the next sheet. Once that one is cutting, he bands all the pieces he can while the machine is busy and/or pulls the next material.

After all sheets are cut for a job, he runs all the parts that have router operations on the back sides (operations not cut in the nest). If the operator is not done with the banding at this point, we might send an apprentice to help with the banding and final sorting if needed.

As for sorting by cabinet at the machine, I don't think any operator could keep the machine running efficiently while doing this. We tried, but gave up. If he does have time to sort, it's by size or type. Bottom line, if your CV is set up right, this system works quite well and can easily run up to 50+/- sheets in an 8 hour shift. More if you have room for infeed and outfeed tables running on two spoilboards.

From contributor M:
A lot of great information here! One thing we often do for customers is assist them with operational efficiency. We follow a Lean 6 Sigma format, which is essentially the Toyota Production System. Often people are uncertain if this is appropriate for their business. "You have to understand, we are different," is a phrase I wish I had a dime for every time I've heard it. The general consensus is that stuff may work for a company making millions of automobiles, but how can it help us? And what is this "lean" stuff anyway?

Lean is simple; Lean is a tool that uses people to identify and solve problems! If deployed correctly, it gets remarkable results. It is difficult to deploy for various reasons and it does take time to do. Lean focuses on waste reduction, and it looks at waste in a somewhat different manner than traditional. It focuses on product flow and the relentless pursuit of excellence in manufacturing.

In order to understand what is going on at your machine, you have to go to the machine and observe! We typically use the terms "value added activity" (what the customer is willing to pay for - changing the function, fit, or form of the raw material into useable goods) and "non value added" (machine set up is an example of non value added - continually striving to reduce machine set up time to keep the machine making the components the customer pays for!).

We do operator balance charts, which is a time study showing how much value added time occurs over a given time versus how much non value added time occurs. We then collaborate with a key team of personnel (including the operators) and restructure how they work. This is not a work harder initiative, it is a work smarter initiative.

It is important to note that we normally advance to this stage after intensive training on lean. Any other method has not proven a record to sustain this or the other dozens of tools Lean/6 Sigma can deploy to affect your bottom line.

Think about what problems you have. If your company has no problems, Lean may not be for you. If yours does have problems, Lean may be for you. Often some problems fall under the Nike "Just Do It" category!

From contributor L:
I'm relatively new to the CNC part of the business, but I've used a principle over the years that simplifies things for me. If I have a feeder on a machine, be it saw, shaper or whatever, the intent is to keep the working end of that machine on the wood. That means the operator must lay out his work so that the machine is cutting product as near constantly as possible. In the case of CNC, sorting does seem to be a bottleneck, but I have found that my old urge to keep that machine cutting something as constantly as possible has led me to examine my material handling on and off the machine.

It sounds like your operation is a good deal larger than mine, and perhaps I'm missing something in your post, but I find it hard to believe that you do not have a push bar and sorting table set up so the machine is running while the sorting is taking place. My priority when my router is running is to keep it running, so as soon as the cycle is over, the parts are shoved off the machine (this is done with a push stick that has a 4' wide end to push the entire cut sheet off the router), the machine is cleaned (I'd love to have this step automated as well, but have a couple of constraints with my router that won't allow it right now), reloaded, and new file started. Then the parts are labeled (marked, in my case, since I like to see the info on the end of my parts on the rack), edgebanded if needed, end or edge drilled for confirmats, and sorted on a rolling rack that will hold the entire job of parts.

These three steps form a 3 sided cell. My priority off the router is to:
1. Mark or label on the sorting table.
2. Edgeband.
3. Organize according to cabinet number off the edgebander onto the rolling rack (this is placed so that I pass it on the way back to feed another piece into the bander).
4. End and edge drill for confirmats. This activity usually takes place as time filler when the first three are complete, if there is any time before the router has finished.

Your router time is probably faster than mine, but my average cycle is about 20 minutes per sheet working by myself. I optimize my small parts first because they take less time on the router and usually have to be EB and drilled, so while the end panels are being milled, which takes longer on the router, I can drill those parts that haven't been done yet. All of this happens within a 25'x35' space that includes the assembly area.

I think you will find that one man on the router is all you need, and if the router approaches the 50 sheet mark as stated above, that is all he will be able to do, and he won't have time to do all that other stuff I mentioned above. You also may optimize per cabinet. That way, the assemblers have something coming at them every 10 minutes or so, right off the bat. I think of my tools as employees. What can I do to keep them producing value added goods through my shop with the least amount of wasted energy, time, material, and non-value-added activity, such as watching the router cut the next sheet?

My first inclination would be to time the router's on versus off time. I built my sorting table (it is on casters as well) for about $75, made a push stick with a 2x2, and cut my cycle time almost in half. Routers are very expensive sorting tables any way you look at it.

From contributor K:

We utilize one man to load and unload our nested base router. We utilize double spoilboards, which allows a very fast load/unload cycle. While the router is machining one panel, the operator unloads and labels the parts from the previous panel and places them on a cart that is specifically designed to handle case parts. Once the parts are unloaded, the empty spoilboard is vacuumed, blown off, and the next panel is loaded onto the waiting spoilboard. We average about 5 minutes per sheet for the machining time, which allows ample time for the previously described operations and still leaves the operator waiting for the machine to complete the machining cycle on the next sheet. Our load/unload cycle takes approximately 1 minute.

Below are several pictures of our case carts. We sort the parts by type, rather than individual cabinet number at this point. We place the sides and shelves on the top layer of the cart, and the tops, bottoms, and miscellaneous parts on the bottom layer. If we had the necessary space, we would utilize roller conveyors to store and transport the parts.

It is no more difficult or time consuming to pull the individual parts off of the case cart once they reach the assembly department than it is to sort them by cabinet at the router. Because the parts are sorted by type and labeled on the exposed edge, finding the needed part is very quick.

This photo shows the adjustable rods used to separate different types of parts. The rods are 1/2" emt.

From the original questioner:
Contributor L, thanks for all the advice. I can see the benefit of a push stick, as well as some of your other ideas.

And contributor K, thanks for the pictures of your cart. We have been going around and around about how to build a cart the last while, and this seems to be the best yet.

From contributor L:
Contributor K is correct about the sorting. I stated that I sort by cabinet, but actually it is by part type, although the smaller parts are grouped with the same cab number. I have a similar cart as contributor K, but I used a Gorilla Rack from Sam's and mounted it on a 3/4 ply platform with casters. Cost, without labor, about $175.00. If you build your own, use very heavy duty casters, 5" or 6" (I got mine from Grizzly), and mount stationary ones in the middle and one swivel type on each end.

My plywood is about 2' from the end, and simply slides onto the router with minimal lifting or positioning. I placed two aluminum angles 3/4" or so wider than the ply to help guide the work both on and off the machine. The offload-sorting table is light 2x4 frame with melamine top and has angle rails as well. These keep the smaller pieces from falling off the table when off-loading the router. It is on casters, so it can be placed anywhere that is best for saving steps or getting it out of the way.

From contributor J:
2 guys is definitely 1 too many. Using 2 spoil boards is the way to go. Youíre removing parts, cleaning and reloading one sheet while another sheet is on the machine. This more than doubles the number of sheets you can cut and makes the second guy even more redundant. There are problems with this method and not all machine owners can fully benefit from this (if you have a fixed spoil board, a gasketed spoil board, or donít have reverse air). Our machine (name not mentionable) has a flat phenolic table with reverse air, the spoil board has no gasket, so for us this method is perfect. We have a stationary table with rollers (under a vacuum lift) at the end of the machine for staging and a mobile cart with rollers to remove the finished sheet.

As for sorting, if itís a small job, a few boxes or a single wall unit, then the parts sort themselves quickly. If itís a large job, say 100 sheets of material or more, we just stack on carts by like parts and send them to the bander. When all parts are edge banded, we sort them out. Nesting, by its nature, shuffles the parts and makes sorting a hassle; if they are all in front of you and are banded, then itís fast and easy to sort. As for parts programmed at the machine, to me this is a thing of the past. All parts should come from the office. I ask our operator to mark any spoiled parts (as he has a paper printout sheet for every sheet of melamine he has to cut) and give them to me at the end of the run, or whenever it suits him. Then I nest them all on a single sheet or however many it takes. Using our software this takes just a couple of minutes. I suppose if youíre trying to start assembly before the entire job is cut youíd run into problems, but if you want the maximum from nesting, then you need to take the right approach to production.

From contributor S:
We have had the same CNC for about 5 years. My advice... One man = 50 cabinets, two men = 150 cabinets a week.

From contributor A:
I think you are wasting your time and money when using any more than 0 operators. In the global economy you should look at setting up robotic loaders/unloaders and automate your facility as much as possible. You purchased a CNC router, which is a robot. Why did you stop there? Take the next step and get rid of the human factor and all the problems that go with it.

From the original questioner:
Robots = too many dollars, unless you would like to become an investor.

From contributor R:
People had the same attitude toward CNC routing 15-20 years ago as you have now toward robotic loaders/unloaders. But now even the guy making guitar picks in his closet is using a CNC machine. Everyone has a CNC router today. The next step is to address the operations before and after the CNC router has done its job.

From contributor E:
Contributor K, did you purchase this cart?

From contributor K:
We manufactured the carts in house. They are 2' deep, 6' long, and 6' tall, not including the casters. There are 2 6" fixed casters mounted in the center, and 2 6" swivel casters mounted at each end. The center fixed casters have a 1/4" thick plate welded between the cart frame and the caster body. This creates a teeter totter effect that allows the carts to turn on a dime.

From contributor C:
Your cart looks like the perfect cart for CNCed cabinet parts. Not being a metal worker myself, I have no idea what a cart like this would cost. Did you have to drill the holes in the square tube stock, or can stock like that be purchased?

From contributor K:
The material cost for the case carts is approximately $200 each. The six casters are about $100 including freight from Grizzly. You will need 50' of 14 ga 1 1/2" sq tubing at approx $1.00 a LF, and 3 sheets of 3/4 particleboard. And several sticks of 1/2" EMT for the adjustable rods.

Regarding the holes for the rods, we drilled them using a standard drill press and a 3/4" twist drill from ACE hardware. The holes are on one face of the tubing only, and are 4 5/16" on center. We are by no means metal fabricators, but are able to weld to save our lives. I don't think you can purchase the tubing with the hole pre-punched in the appropriate sizes and locations.

We have approximately two dozen carts fabricated in this manner in various design configurations, each type for a specific purpose. The actual labor to fabricate the carts is the expensive part. However, having exactly what you need, engineered to do your job perfectly, is priceless.

From contributor C:
Thanks. Does not seem that bad to me. Now to place the idea in front of the boss.

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

Comment from contributor A:
I currently program and run all sorts of custom parts on three different CNC machines - a 5 by 10 CAMaster with ATC, a 48 by 96 ShopBot, and a 10 Spindle CNC 4th axis carving machine. I load and unload all my parts and design and program all my vector art and respective code for these machines. One guy equals less errors and less payroll. Three machines equal lots of consistent parts time and time again. Best of all, if something is screwed up, I know who did it!