Interviewing CMM Programmer

[jerrykittle]

My company is seeking a new CMM programmer. Since I'm not a programmer my self what would be some questions a some sort of a test that I could give someone to see if they really had the experience we were looking for?

We use PCDMIS and we have a Brown and Sharpe 775 model.


Jerry

[Don Ruggieri]

Try this. . . .

what would be some questions a some sort of a test that I could give someone to see if they really had the experience we were looking for?

We use PCDMIS and we have a Brown and Sharpe 775 model.


Jerry

Ask him how to use Passthrough planes

[jerrykittle]

Don,

Is there an attachment you posted; if so I'm not able to open it for some reason?


Jerry

[J. Pick]

If you have the CAD or CAD++ version of PCDMIS, you need to know if your applicant has experience with CAD models. Inversely, if you do not have the CAD version of PCDMIS (PCDMIS Pro), your new hire should know how to program without the crutch of a CAD model (from a drawing/print). Ask if they have been through Brown and Sharpe's training courses. Personally, I would want someone who could also work off of a surface plate as well, but maybe you just need a dedicated CMM programmer. Also, make sure that they have programming experience, and not just operating experience. There is a major chasm between the two. If your applicant has experience programming CMM's using a different software, know that there will be a steep learning curve and that you will need to send that person to Brown and Sharpe training to get the full benefit of their experience (check out the prices on this website).

[craiger_ny]

Hire the minority, unless a hot chick applies.

[#2#]

Jerry,
Are you hiring a "new" CMM programmer as someone has quit and you need to replace that person?

Or, are you hiring a "new" CMM programmer as you have just purchased a machine and now need someone to run it?

If the first case, did your current programmer leave yet? Is there any way they could sit in on interviews?

In either case, you should either look for proof (certificate) they have been to PC-DMIS training or be willing to send them to training right away. It might not be a bad idea for you to go to at least the basic class and then you will have some idea of how it works and what to look for in a programmer.

Also, in either case, it would be wise to look for someone with traditional (plate) metrology experience. A CMM is just one more tool used to inspect parts. If you have no idea how to inspect a part with a caliper, a cmm isn't going to do you any good.

As far as a test... You could give them a simple part that can be measured with calipers, micrometers, etc. and ask them to write you a quick program to measure a few features. If they can give you results within a few minutes, they probably know what they are doing. Also, you could check with the hand tools and see if their results are reasonably accurate.

Good points above on CAD vs. non-CAD.

[bw_bob]

1st thing is ask if he,she knows what CMM stands for, what I,J,K vectors are & what are X,Y,Z coordinates. If they know that like craig said hire the hot chick.

[layout tech]

Don

I have been programming for 5 years and with usually no CAD data and
complicated parts. I consider myself a fairly good programmer. I have done outside programming for other industry in our area for the last 4 years on contract basis. I have yet to use a passthrough plane I use Move Points.
Asking me to explain passthrough planes I might fail. I understand the concept but have never tried to use.

I might not even qualify using that criteria. Pretty restrictive.

[Peter Warcholyk]

layout tech,

I concurr, I have not used pass through planes. I also utilize move points.

Vector direction, CMM, manual layouts on a surface plate, etc., important questions to ask.

Regards,
ZydecoPete

[Jan d.]

I think you need someone with a brain on his/her shoulders. If they're smart, they'll figure it out.

I have been burned by this before. Someone that could dazzle me with operating a software package, but had no clue as to what he was really doing. I'd look for the individual showing the most aptitude, whatever that may be. If they exist upstairs, they will pick up any software packages rather quickly. You'll have a more loyal employee that way too...

Ask questions to do with basic math. Probe their trig skills and whether they understand some basic stats. Then pull out a print with all kinds of GD&T on it and just ask that individual to talk about that print. Sit back and listen. See whether they can quickly zoom into FCF's that are hard to measure. Just have them talk out-loud about what they see. The individual that is the most to the point on a print, will probably be a front runner.



Jan.

[#2#]

I agree with what Jan and layout tech are saying. A trainable employee is much more valuable than one who will never learn anything new.

[cmmguy]

Don

I have been programming for 5 years and with usually no CAD data and
complicated parts. I consider myself a fairly good programmer. I have done outside programming for other industry in our area for the last 4 years on contract basis. I have yet to use a passthrough plane I use Move Points.
Asking me to explain passthrough planes I might fail. I understand the concept but have never tried to use.

I might not even qualify using that criteria. Pretty restrictive.

Having read most of Don's posts, I would guess he was being sarcastic.

[kbotta]

I think you need someone with a brain on his/her shoulders. If they're smart, they'll figure it out.

I have been burned by this before. Someone that could dazzle me with operating a software package, but had no clue as to what he was really doing. I'd look for the individual showing the most aptitude, whatever that may be. If they exist upstairs, they will pick up any software packages rather quickly. You'll have a more loyal employee that way too...

Ask questions to do with basic math. Probe their trig skills and whether they understand some basic stats. Then pull out a print with all kinds of GD&T on it and just ask that individual to talk about that print. Sit back and listen. See whether they can quickly zoom into FCF's that are hard to measure. Just have them talk out-loud about what they see. The individual that is the most to the point on a print, will probably be a front runner.



Jan.

LOL Jan - that could back fire! As...oh! Look at all of these RFS DRF's...VERY complicated for a CMM to construct TGC's of constrained and located datum features...
Or.... WOW! Look ! Someone added a material condition modifier to a planar surface! cool!!! (and then bolt for the door... )
(disclaimer: Not to offend anyone working with MMC on planar surfaces... )
kb

[afenner]

I've been programming for 3 years and have never used a passthrough plane.

[ag162]

Hello,
Whether it is a passthrough plane or a clear plane, knowing the language is important. Ask the questions regarding GD&T, how would he or she set the part up and how would they do the alignment. Can they calculate TP? Can they calculate bonus tol? Of course, there is the bonus question to see if they can think. "What can you do for a program in which the styli misses the part during a program, causing the program to stop running"?

VP A.Gore