Re: How to carry out Gage R&R for automatic machines?
But our case is,Its a automatic measuring line system.there is no operator to place the part which we are measure.parts (we measure camshaft,crank shaft etc) comes through a conveyor system and it stops on the measuring station a lifter is lift the part for measuring.the meadsuring system measure the diameter, height ..........,we use digital sensors and mini probe some times air gauge then the parts goes to the next measuring station like that the process going.
For this system how can i study GAGE R&R.
Can you find the original MSA from when the system was bought and commissioned, there should be one somewhere? It should give you lots of clues how to proceed.
You likely already have hundreds or thousands of measurements. Why not just plug them into software and see what you get.
If you get a really bad R&R read more here. 10 parts in a row from your process likely have very little part to part variation (in fact I am certain they would given they are crankshafts). This creates a poor R&R result.
If that happens , select 10 parts that show normal expected process variation (multiple lines, different set ups, different shifts, different weeks, different grinding wheels, etc.). Or you may have to make special parts for the study.
If you need to get good R&R numbers just to satisfy a customer, get 10 parts ranging from under spec thru over spec. It should pass. Very few customers know the problem with this method. Or even better use 2 each of 5 different parts. A tiny crank all the way to a monster if you have them. You will get a great R&R. A totally bogus study but again most customers know only to look for less than 10% R&R.
If you are doing a study to understand and improve, please ignore the above, and good for you (and are you hiring?)
But, hey, why not just grab the data and get started.
The people on this forum are genuine experts, I am just a user. They will help.