And Then We Decided To Automate, Part III
The machine’s been in the shop for more than six months now. I’ve read articles and heard many arguments for and against CNC technology, but I can assure you the only regret I have is that I didn’t purchase one sooner.
I’ve often heard people lament the much over-hyped learning curve. “Mark,” somebody would tell me, “it’s going to take you a couple of months and a bunch of broken countertops before you get the thing up and running”
Yeah, whatever. That may be the case with bad machines running lame software. Or, it could be the case if you are using pins to position parts and can’t get the hang of it. But, getting up to speed hasn’t been a problem here.
It’s difficult for me to comment on other CNC machines, as my knowledge comes from running our Northwood 138. I’m certain that some of the high-end CNCs enjoy the same benefits.
On the third program that I wrote, I did something utterly stupid. The end result was the finger bit coming down on the corner of a top and causing a mini-crash. (I picked the corner up, glued it back on, top-polished it and drove on.)
About three months ago, I had some visitors in my shop and was yakking while programming and showing them how you could “strip” tools paths and other wonderful things. And, I forgot to reinsert a command, which caused a crash and broke a top.
In both cases, the crashes were not a result of bad training or a learning-curve issue. The truth is, I wasn’t paying attention.
I’ve since written and executed hundreds of programs without incident. The software that comes with my machine has two simulation modes (or sims) that allows you to check for tool crashes. The 3D modeling allows you to actually look at the finished part in 3D; you can rotate it, check the edge work and corners, etc.
If you follow your procedures, runs sims on your programs, and don’t allow yourself to be distracted while programming, you’ll have no problems with the infamous learning curve. If you’re a clown, join the circus – don’t buy a CNC.
What you do learn over time is speed. What took me 40 minutes to program and check during the first month now takes five or 10 minutes.
You’ll also gain confidence. On many simple programs, I don’t run simulations at all anymore.
You’ll also figure out how to batch your jobs for maximum efficiency. The simple things you don’t learn in school, like how to organize your vacuum pods and ways to keep the laser calibrated, become self-evident.
No matter what machine you wind up purchasing, I would give serious consideration to these technologies:
• Fanuc controls;
• Overhead laser;
• Bar-code scanner system; and
• Laser tool measurement.
The Northwood 138 comes with an ingenious bar code system/scanner. We use it two different ways.
The main use is for scanning in the programs. When I’m on schedule, I digitize my templates and write the programs the day before putting the stone on the machine. Each batch gets transmitted wirelessly from my programming computer to the CNC; the computer prints out a page that shows the batch name, program number and a picture of the CNC bed with parts and pods placements.
What makes this ingenious is that ANYONE in the shop can then run the CNC. It takes four simple steps.
1. Take the bar-code readers and scan the printout with the batch information. This does two things at once; the program with the g-code for running the CNC is loaded into the Fanuc controls, and the overhead laser system activates to instantly projects a beam and show the placement of vacuum pods.
2. Once the pods are in place and the “lock pods” button is pushed, the overhead laser shows the exact position of the parts.
3. Line up the parts to the laser, check to make sure everything looks correct, and push the “lock parts” button.
4. Push “cycle start.” And, at that moment, the CNC’s working and the job is in production.
It still strikes me as ironic that, with this arrangement, the lowest-paid guy in my shop sets up and runs the CNC.
We also use the bar-code scanner for loading mini-programs. I wrote one for machine warm-up, replacing the MDI (manual data input) series of commands to get the machine ready for the day.
Now, when my lowest-paid guy gets to the shop ahead of me, he powers up the CNC and scans the bar code of that mini-program. The machine warms up – after which he scans the first batch of the day, hits cycle start, and makes me money while I’m grabbing coffee at Starbucks.
The other major feature of our CNC is the laser tool-measurement device, which is something I didn’t really appreciate at first. Back in the “old” days of CNC, you had to periodically measure your tool and update the machine. This was time-consuming and prone to human error.
Now, if I want to update a profile set, I simply scan the bar code for that set and push cycle start. The machine automatically grabs the tool, measures it, puts it back, grabs the next one, and continues until the entire set is completed. Then, it updates the wear-vector software and adjusts the compensation factor. In other words … something that used to be a pain in the butt is much simpler to do.
When you fire up your CNC, you will start to realize that processing stone with a CNC is like watching grass grow. When we fabricate, we do several extra steps that makes the grass grow even slower.
Firs, we calibrate our slabs. We run a special tool that makes the edgework on the slab the same thickness.
We also Z-mill, allowing us to get “perfect” seams. When you use the Z-profile to machine your seams, it grinds back the chips and makes the edges perfectly straight – supposedly, .0003 of an inch perfect. (I will dedicate an entire article to building perfect seams in the near future)
These two steps add about 20 percent to the time it takes to process an average kitchen. Many shops running CNCs skip these steps; however, we’re a high-end custom operation and don’t mind the trade-off. We purchased the machine to perfect our work, not more kitchens.
We don’t use the CNC to polish stone; we only run the metal bonds and then finish by hand. No CNC in the world can match a hand finish on standard edges. We do run polishing wheels on the more-exotic edges, such as a triple pencil.
The machine increased our profits right out of the gate. Rather then build and install more kitchens, we went in another direction; we only install one big exotic kitchen per day, along with vanities, fireplaces, and other custom work.
We eliminated two employees and almost all of our overtime. This alone is making the machine payment.
The day the machine showed up I told our operations manager to raise prices 10 percent. I reasoned that when our customers come to our shop, they would be impressed by our gear and fork out a little extra dough. This proved to be correct.
Every part we make is perfect to the template., and I mean perfect! The reveals on the undermounts are consistent and true. The holes are pre-drilled for the spigots. That big radius on the raised bar is actually a true radius, and it’s actually a little weird.
Our installs are going much faster with less overtime. My shop guys are so much happier as the work they have to do now is much less stress on their bodies.
Nature and business, however, abhors a vacuum. As soon as we figured out that we could meet our goals using only 80 percent of our machine time in an eight-hour shift, we started making things for other fabricators.
It started when some buddies of mine, who are just starting out, brought tops over to run undermounts. Then, they started bringing templates. The next thing we knew, we became – eek! – a factory.
Last month, we fabricated four kitchens that we didn’t template or install. I expect we’ll do many more in the future … and we do them with a big grin on our faces.
A good friend of mine in the industry told me this. He said: “Mark, you’re buying a really expensive printer that can only print three or four pages a day. Use it to print 20-dollar bills instead of ones.”
Our company is marching down that path. We never wanted to take over the world – just build perfect kitchens one at a time. It’s changed my life and the lives of our employees, not to mention the quality of our installed work.
If you would like to see more pictures of our CNC being made, set up, and operating, visit the photo gallery at www.stoneadvice.com. If you have a question post it in the forum and our panel of moderators will try to answer your query.
‘Til next time – Mark Lauzon, stonecutter.
Mark Lauzon is a fabricator in Oregon and the administrator of www.stoneadvice.com, a Website dedicated to slab fabrication.
This article first appeared in the November 2005 print edition of Stone Business. ©2005 Western Business Media Inc