Dear All,

I am on a project which includes a sensor to control if there is any reverse label.

Front side of label is red, other side is white. If it sees red, it is OK. If white, it gives warning.

And in each day, I have to control it by putting the label reversely and check it (according to procedure).

But I want to control the sensor automatically if possible without use of reverse label. Could I do it? Or is there any interesting idea to replace instead of this procedure?

Thank you.

Best regards,

 

The daily test sounds like a standard QA procedure. You are checking the complete system (not just the sensor) to make sure that bad parts get rejected. Most companies would also keep a record or check-sheet to show that the test was performed. You might find a faster and easier way to perform the check, but you can't realistically eliminate it. It's not just the sensor that might go wrong.

In most companies the machine operators or QA personnel would do this as part of their start up checks at the beginning of their shift.

 

That's exactly same as ours...all you said!! ok, is there any method to reduce business of technicians or decrease risk due to control by 'human'?

 

There is no general answer to this, as it depends on the nature of the manufacturing process and the design of the machine. When you are purchasing a machine, this is one of the important characteristics that have to be evaluated.

It also depends on whether it is an in-process check (with no final effect on the customer) or part of the final test procedures (it helps determine what gets shipped to the customer).

I have been involved with a lot of bar code label applications, and it's only what ended up on the part that really mattered. We would apply the label and then read it (that's the only way to know that it is on the part and is readable).

The usual test for that was to remove a label from the roll and replace it with a different one (with a different bar code). This would normally be done when the roll was being changed, so the person changing the roll would be already handling it anyway. We would just remove a label from a spot a dozen or so labels from the start of the new roll and replace it with one meant for a different model. This meant the test was done on a per-roll (or per model) basis, rather than on a timed basis. As I said before, you have to evaluate things case by case.

 

Look at process sequence to see if test can be inserted in-line. Following is an example:

Problem:
Check calibration of Film-Thickness gauge on a cellophane-web machine.

Solution:
Extended lateral travel of sensing-head to include fixed thickness samples, mounted on each side of web!

Phil Corso