How do you estimate / quote programming times?

I don't have any answer, but the same principle as for "five minutes per touch" should apply, note how long it takes you to create programs, tabulate this against feature count, dimension count, touch point count, number of tip angles, whatever... and try to find some usable correlation.

The results will probably differ wildly from person to person...

Guess based on your own experience and then double it!

Things could change on how complex the part is as well. Taking datum targets and alignments into consideration. Some parts take minutes to program, others take days to program. I've seen everything from Cessna engine casings with over 2500 dimensions on 6 pages of prints to 2 dimensions on 1 page print. One takes weeks the other takes minutes.

On-line or off ? DO you program offline from an assembly ( part on fixture ? ) or then need to possibly heavily edit when you get your part and then build your setup ? More time up front that way but 5x less time at prove out.

If you are working with engineers, however long they think it should take you then multiply that by 5.

I do all my programming offline and I always try to have a really fleshed out idea of the set-up before hand. I try to spend all my time up front, so when we get the actual part in our hands, we don't spend too much time on the proveout.

Typically, I will go through a print, however many pages, and highlight what I think can go on the CMM and what can't, and what I will try to get, and all the datum features. From there, I can usually make a pretty good guess how long it will take, (then I double it! ;-) ) But, I guess AndersI has the right idea.

"Note how long it takes you to create programs, tabulate this against feature count, dimension count, touch point count, number of tip angles, whatever... and try to find some usable correlation.

The results will probably differ wildly from person to person..."

The trick is, my boss or someone ELSE would like to look at the feature count or dimension count, and come up with a ballpark programming estimate.

Thanks!

Guess based on your own experience and then double it!

...twice!

The trick is, my boss or someone ELSE would like to look at the feature count or dimension count, and come up with a ballpark programming estimate.

Sounds like a stick to beat you with

Sounds like a stick to beat you with

... yeah, this ...

"It'll take 1 hour to replace your brakes, sir."
"Well my cousin Billy Bob says it shouldn't take more than 15 minutes!"
"Is Billy Bob a mechanic?"
"Well, no, but..."

a general rule for me is 5-10m per dim. Normally if you are a beginner and not familiar with the parts/product 10min is ok. If you are familiar with the product I would say 5-7 min per dim is ok.

example: 10m per dim
Say you have a TP callout, you would count 2 Basic dim +TP callout 10+10 +10=30m

Say you have a TP callout on 10 holes, you would add BASICs (angle +locations)+ BHP dia+ 10 holes:

10+10+ 100= 120m

I know it may sound like a lot, and it is, that's why its a general rule for unfamiliar parts and for a beginner. Typically I use 7m rule on the familiar product and 10 for unfamiliar, to ensure I have some wiggle room.

Please don't tear up my post, this is just what I use, doesn't mean its for everyone.

I started with using 10m/dim at first then I changed it to 7/m and it worked out well for me.


There are times, when you may have to deal with something you not sure how to do, in that case I would change it to 15m/Dim. There times when things just go bad, and you want to make sure you do not underbid.

How many times have you recreated a new program from scratch, either due to data loss, or due to bad programming? How many times, had it take you a week or more to figure out how to do something? these are the questions you need to ask yourself. and add time accordingly.