hexagon logo

How long does it take you to program?

Let’s say someone gives you a metal part with 150 dimensions. How long would it take you to study the print, figure out fixturing, create a setup sheet, and program it? You’ve never seen this part before and it’s somewhat complex. Assume the print makes complete sense to you after studying the print—so you don’t need to ask the designer any questions.

Also, would the program run perfectly the first time? If not, how long would “proving out” the program (making adjustments) take you?

I ask these questions because I get them a lot being the only programmer at a significantly large company with 3 machines. I’m curious what other people’s experiences are, and I’m open to any tips. I will state my answers to these questions in one week. Hopefully I get a lot of responses.

  • Wow, thanks for all the replies, everyone! Some people made some great points, like how it takes a few hours to "prove" out the program. Some people in this thread gave some pretty short quotes. I can program 150 dimensions in a few hours if I really tried, but the program would be a flop 20% of the time. Either the alignment would need to be redone, or even worse, I could get to some features and realize they're unreachable with my probes, so I would basically need to redo the program with a different part orientation.

    Here's my answer: I usually quote a full 8 hour shift for any program. Some short programs may take me 3 hours, and some long programs (150+ dimensions) may take me 8-16 hours. I try to give an average of "one work day," because a part with not many dimensions could take a while if it is hard to design fixturing. Also, sometimes I have questions about the print I am given. Either the datum structure is not stable (or makes no sense) or some dimensions don't match the model nominals.

  • There's a lot of good detailed responses here, and I have a follow-up question.

    Is there any tips from the people who have responded to improve programming speed that could help just about everyone? Even if it's a small thing.

  • The only thing I've found is that you can create sets and constructed features using auto-feature hitpoints. So like if I wanted to create a plane out of two planes I could do pln1.hit[1..pln1.numhits],pln2.hit[1..pln2.numhits] That would create a plane out of all the hitpoints of pln1 and pln2. Or, if I wanted to dimension the profile of both planes, then I could create a constructed feature set in the same manner. Instead of creating constructed feature sets of a million auto points and having to select them all individually.

  • I recently had to program a part that is less than 36" long, less than 3" wide, and less than 1" thick.  Customer (who is clueless) had ballooned up a print with the 'required' dimensions (which was 75% basic dimensions, so that was WRONG WRONG WRONG!) and ended up with 2,313 dimensions (yes, over 2 thousand dimensions on the report).  Takes my Global 3.5 hours to run and it makes a 200 page report.  Took 48 hours to program and prove out (as well as make MY Excel report file).  1.25 minutes per dimension.

  • We do Excel FAI reports too, which is a real pain. I am using Excel Form Report now. But there is some formatting I have to do, like adding all the GD&T characters to the nominal values to make the FAI report look presentable. Like, instead of true position having a nominal of 0, I manually enter in |TP|.005|A|B|C|.

  • my excel form is a 100% custom creation and it is graphical & textual.  I makes a 'pretty picture' for people that can't (or won't) understand numbers.  I makes a textual summary, it makes a raw data report, and it makes a Pcdmis-type report as well.  The graphical looks like this:

    Yes, it can take some time to set up.  Screen captures from Pcdmis showing the points (dots), then you have to drag the markup cells to where they 'fit' then add leader lines.  Yes, takes a bit of time, but once you check the parts, it takes less than 2 minutes to have a report printing out, and there is no way for someone to 'mis-understand' what is good or bad anywhere on the part.  The borders of the cells turn RED when they are OOT

  • Depends.   150 dimensions on a trade show type of part, like the demo block?  An hour or two.

    Extremely critical parts where all kinds of tricks are needed?  Could take weeks.

  • Interesting. Maybe I should consider something like this. Because, engineering is constantly changing balloon numbers. They refuse to manually balloon the prints, so sometimes all the balloon numbers change at a small rev update. With something like this, I suppose I could leave out balloon numbers (since the reader knows where the dimensions are located).

  • my 'master' Excel file is all 'set up' and ready for pictures & data to be put in.  It can handle UP TO:

    270 SPC features

    600 surface features

    480 trim features

    80 hole LOCATION (not position) features

    80 slot LOCATION (not position) features

    120 hole POSITION featues

    80 slot POSITION features

    480 'linear' dimension features

    It is in a 'constant' updating process and is currently over 144MB in size.

    Once it is copied and filled in a for a job, all the extra (un-used) stuff gets deleted and some stuff gets moved from tab to tab so all I have to do it copy & paste the data from a DatapageRT data report into it and then print it.

    Quite a bit of the time spent on that 'stupid' job was taking that master file and making the 'linear' section 6 times as big to hold their data.

  • Most of us work in an 8 hour day too, depending on your habitat you can subtract 30-45 minutes for lunch and breaks. Seemingly never a "full day." 

    We loose 10-30 minutes every morning for a "morning meeting" another 40 minutes for lunch and breaks we are required to take. We typically have to attend to other people's needs and are maybe left with 6 hours to program if we are lucky, many times less than that.