Best Attribute Data Sampling Plan - Early Production Containment

[beaser3]

Hi,

My company has recently implemented a final inspection for the first three production runs of a new product. We are currently doing 100% visual inspection on these runs. The process has been successful in finding defects but it is also very labor intensive considering some of the runs may be 50,000 pieces or above. We would like to implement the use of sampling plans for the inspection, possibly using a tightened, normal and reduced inspection. Can anyone tell me which sampling plan would be the best to use for attribute data? By the way, many of our parts are automotive. Thanks in advance for any thoughts/suggestions.

 

[sowmya]

Check IS 2500, attribute sampling plan. You need to choose AQL based on your requirement

 

[Kales Veggie]

Hi,

My company has recently implemented a final inspection for the first three production runs of a new product. We are currently doing 100% visual inspection on these runs. The process has been successful in finding defects but it is also very labor intensive considering some of the runs may be 50,000 pieces or above. We would like to implement the use of sampling plans for the inspection, possibly using a tightened, normal and reduced inspection. Can anyone tell me which sampling plan would be the best to use for attribute data? By the way, many of our parts are automotive. Thanks in advance for any thoughts/suggestions.

My two cents: If you are finding defects, you cannot reduce your inspection. The process must be improved. To reduce inspection, you must have evidence.

If your customer find a defect (and they expect close to 0 ppm), it will be more costly than 100% inspection.

My suggestion is to improve the process and then have data to support reducing inspection.

Also you are finding defects, you should review your PFMEA and adjust the Occurrence accordingly.

Also the defects may be time / machine restart / shift change / material related (say new batch, morning start-up). There may be ways to determine when the defects occur and adjust inspection frequency.

Simply going to an reduced inspection frequency is very risky.