Shall we raise SCAR/CAR for every defect as per our current procedure?

andreww · Sep 26, 2012

We have a procedure on controlling of non-conformance product. As per procedure, for every defect product, QA shall raise defect tag, book defect tag in database and raise SCAR to supplier or internal CAR, then we need to do RCA, CAPA, follow up to close it off. It means for every defect, we need raise CAR/CAPA. Is it practical?

For example, for one time occurred defect in 12 months (cost only $10), short delivery (supplier agree to send missing parts, only one time occurred in 12 months), internal defect (only quick fix) and etc. I don' think it is necessary to raise SCAR/CAR for these non-conformance.

Shall we have "filtering'' system? I.e. all defects shall be booked in database and we monitor them (control chart, Pareto chat and etc.), only critical one to QMS, we need to raise CAR/SCAR. If this is the way, shall we have to define which is critical to QMS in procedure? Please advise. Thanks.

Michael_M · Sep 26, 2012

You may be able to add something that adds a level of traceablity. Corrective action and Non-conformance. You can then change your procedure to state that Non-conformance's are tracked but not necessarily acted on. This would give you history of what happens without over doing the corrective actions. However, I believe you would then need to specify what is a non-conformance and what is a corrective action as well as what moves a non-conformance into a corrective action.

andreww · Sep 26, 2012

Thank you very much. Michael. Good suggestion!

Mikishots · Sep 27, 2012

andreww said:
We have a procedure on controlling of non-conformance product. As per procedure, for every defect product, QA shall raise defect tag, book defect tag in database and raise SCAR to supplier or internal CAR, then we need to do RCA, CAPA, follow up to close it off. It means for every defect, we need raise CAR/CAPA. Is it practical?

For example, for one time occurred defect in 12 months (cost only $10), short delivery (supplier agree to send missing parts, only one time occurred in 12 months), internal defect (only quick fix) and etc. I don' think it is necessary to raise SCAR/CAR for these non-conformance.

Shall we have "filtering'' system? I.e. all defects shall be booked in database and we monitor them (control chart, Pareto chat and etc.), only critical one to QMS, we need to raise CAR/SCAR. If this is the way, shall we have to define which is critical to QMS in procedure? Please advise. Thanks.

You could technically do it that way, but it's not practical. Say you had ten parts each produced a week apart, and each part had the same defect. What benefit would ten individual CARs have over one that simply states the defect has occurred more than once and needs to be investigated? This is the balance between NCR and CAR processes.

Monitor and measure your process, track the results, perform data analysis and implement preventive actions based on the results.

andreww · Sep 27, 2012

Thank you. Mikishots. You are right. If these ten parts have the same defect each week, I will raise in one CAR that simply states the defect has occurred more than once and needs to be investigated. The purpose of using defect database is to have data pool to monitor, control and action on this data. After action (CAPA), I check the effectiveness of CAPA by looking into data trend and vital few again. (PDCA cycle).

Shall we raise SCAR/CAR for every defect as per our current procedure?

andreww

Involved In Discussions

Michael_M

andreww

Involved In Discussions

Mikishots

andreww

Involved In Discussions

Similar threads