hexagon logo

CMM Operator Training

How do you train your operators? Any advise is welcome. Don't be a turd.

We have been intermittently slow at work so I have been building basically a curriculum to train new operators (we are expecting to hire people at the beginning of the year).

My "Curriculum" is basically just notes on how to run the CMM, but with some pictures and a few step by step guides. I have used one of my hand mic guys as a path finder/Guiney pig and it seems like every time I get done with a training section there are so many things that I have left out that he immediately stumbles over.

Idk I am pretty good at explaining things in person but putting them down on paper is a bit harder.
  • All my programs operate the same way: 1) Execute Program 2) Setup Instructions 3) Form to fill out 4) Program runs 5) Report

    The basics are needed: Where's the PC-DMIS icon, how to open a file, how to know which program to run, cleaning the part before running, importance of fixturing the part correctly, how to read the report, etc.
  • All my programs operate the same way: 1) Execute Program 2) Setup Instructions 3) Form to fill out 4) Program runs 5) Report

    The basics are needed: Where's the PC-DMIS icon, how to open a file, how to know which program to run, cleaning the part before running, importance of fixturing the part correctly, how to read the report, etc.


    I have a very standard procedure as well and I even build all my setups in the CAD assembly so the operator has a graphical representation of the setup.

    Yeah but how do you calibrate a probe? How do you know what probes are on the machine or in the probe rack? How do you build a probe and add it to the probe rack information?
    When dimension 19 has 4 holes and one of them fail on the true position how do you tell the machinist which hole is failing and in what direction?

    This is what I am trying to cover. Basic notes so that I don't get phone calls while I am on vacation.
  • I have done lots of training in a multitude of things. I found that a Show, Watch, Train Me method worked the best. I would show an operator a basic task and answer any questions. Then, I would watch them do it, and answer any questions. Finally, once I felt that they had a good grasp on it, I would have them show me how to do the task, like I had never done it, and ask them questions. It often worked quite well.

    After explaining some of the basics, I would have them do a 3-2-1 alignment on and measure a 1-2-3 block. After that, I would have them write a DCC program on top of their manual measurements.

    My old shop wanted a work instruction for CMM programming. I kept telling them that a CMM programming work instruction would look like an encyclopedia set. I have setup instructions and pictures for programs I have done but that is about it.
  • My old shop wanted a work instruction for CMM programming. I kept telling them that a CMM programming work instruction would look like an encyclopedia set. I have setup instructions and pictures for programs I have done but that is about it.


    This is basically what my shop would like to have as well. I have told them the same thing but I didn't see any harm in at least doing a few of the operator functions. Right now I have 10 "modules".
    Some of them are just lists and some are a bit more detailed. I think the longest one is 3 pages in word.

    File Locations.
    Machine Operation (power on, jog box).
    Navigating PCDMIS.
    Filling out header information (I have a form that they fill out).
    Building probes (how to measure a probe, probe rack settings).
    Calibration.
    Machine setup.
    Execute program (manual hits and read points).
    Read CMM data (diagnose issues).
    Output CMM data (Excel settings).


  • This is basically what my shop would like to have as well. I have told them the same thing but I didn't see any harm in at least doing a few of the operator functions. Right now I have 10 "modules".
    Some of them are just lists and some are a bit more detailed. I think the longest one is 3 pages in word.

    File Locations.
    Machine Operation (power on, jog box).
    Navigating PCDMIS.
    Filling out header information (I have a form that they fill out).
    Building probes (how to measure a probe, probe rack settings).
    Calibration.
    Machine setup.
    Execute program (manual hits and read points).
    Read CMM data (diagnose issues).
    Output CMM data (Excel settings).


    Those all seem fairly easy. I actually think I ran across a step by step instruction on the forum for calibration that you could almost cut and paste. I was asked to write work instructions for writing inspection programs.

    I would put a disclaimer in every one that says: If you are not sure about something, ask. I have spent hours trying to fix something that someone took a guess on.


  • Those all seem fairly easy. I actually think I ran across a step by step instruction on the forum for calibration that you could almost cut and paste. I was asked to write work instructions for writing inspection programs.

    I would put a disclaimer in every one that says: If you are not sure about something, ask. I have spent hours trying to fix something that someone took a guess on.


    I did! I also put, at the very beginning of each one, to read the entire thing before starting because my inspection guy would ask me questions that I addressed later down the page. The thing that I have battles the most is getting someone to actually read the dang thing before asking 1000 questions.
  • This is something that I am struggling with currently. The lady that runs the CMM on the off shift is over 60 so technology is not her strong suit. She just got a smart phone a few months ago. I spent 45 minutes on Monday on the phone walking her through a manual setup. (3 points, one circle with sample hits and another single point) I was hopeful that a simple touch the part where the red dot on CAD is would suffice..... It did not.

    I have also been trying to train them from time to time but its never on a consistent basis so its always just recapping what we have already covered. Fortunately most programs/parts don't have any issues. But when they do....AngryAngryAngry

    I also have to keep training event logs for each session and document what we covered and how long I trained said individual. Which I do not agree with because some people will take that and say "well you spent 15 hours on alignments. Why cant you align a part?"

    What I have done though is created a 100 question test that covers many of the basics (Datums, print reading, GD&T, when to use X feature, etc) and some more advanced programming stuff (sample hits, avoidance moves, alignments, Feature Construction, etc) and I have them take the test from time to time to gage what I need to focus on during the limited training time we do have.


  • This is basically what my shop would like to have as well. I have told them the same thing but I didn't see any harm in at least doing a few of the operator functions. Right now I have 10 "modules".
    Some of them are just lists and some are a bit more detailed. I think the longest one is 3 pages in word.

    File Locations.
    Machine Operation (power on, jog box).
    Navigating PCDMIS.
    Filling out header information (I have a form that they fill out).
    Building probes (how to measure a probe, probe rack settings).
    Calibration.
    Machine setup.
    Execute program (manual hits and read points).
    Read CMM data (diagnose issues).
    Output CMM data (Excel settings).


    The only other thing I can think to add is Troubleshooting. How to manually move the probe from an interrupted hit. How to make sure you didn't do any real damage. How to clean the styli without messing things up. How to clear an E-stop if you bump into it. Stuff like that.
  • Yeah my #1 operator right now just turned 65. He is actually very ambitious to learn and has started programming within the last ~3 months. I think he is a pretty rare occurrence.

    I never want to deny someone information so if he wants to and is willing to learn I will train him until he decides to retire but it is defiantly disheartening to know that eventually he will leave and all that training will have been a loss.

    He says he wants to work another 5 years so honestly it is probably for the best. If I had an 18 year old in here to train he would probably move on to a new job within 2 years of training for more money. Maybe it is better this way.
  • This is something that I am struggling with currently. The lady that runs the CMM on the off shift is over 60 so technology is not her strong suit. She just got a smart phone a few months ago. I spent 45 minutes on Monday on the phone walking her through a manual setup. (3 points, one circle with sample hits and another single point) I was hopeful that a simple touch the part where the red dot on CAD is would suffice..... It did not.

    I have also been trying to train them from time to time but its never on a consistent basis so its always just recapping what we have already covered. Fortunately most programs/parts don't have any issues. But when they do....AngryAngryAngry

    I also have to keep training event logs for each session and document what we covered and how long I trained said individual. Which I do not agree with because some people will take that and say "well you spent 15 hours on alignments. Why cant you align a part?"

    What I have done though is created a 100 question test that covers many of the basics (Datums, print reading, GD&T, when to use X feature, etc) and some more advanced programming stuff (sample hits, avoidance moves, alignments, Feature Construction, etc) and I have them take the test from time to time to gage what I need to focus on during the limited training time we do have.


    I tried something similar, with the tests. Problem was, our training system was set up to where a score of less than 80% was a failure nd required corrective actions and such. You either passed an assessment or there was an issue with the training that had to be addressed. We started writing multiple choice questions with obvious answers i.e.:

    What are the key parts of an alignment:
    A) Level, rotate, and origin
    B) Wash, rinse, repeat
    C) Only management is allowed to alignment
    D) All the above

    Best part was, management couldn't figure out how people never knew anything but passed training tests with 100%.