Re: Difference between "Document " and " Record" in ISO 9001 Procedure
Andy's "elegance" and Jim's "trail" are important factors. The point of both men, I think, is that changes to any documents, especially records, should follow a protocol, not willy nilly because some guy "decided" there was an error and erased the original entry in favor of the new one without leaving a trail of the fact a change was made and why it was made. The trail helps prevent silly mistakes as well as intended fraud.
The elegance comes from being able to review the accurate history to help identify opportunities for improvement in processes, designs, material, etc.
An example of how the "trail" helped generate a improvement in an inspection process which eliminated nearly 90% of data entry errors and cut the net inspection time on a product by nearly 70%:
Opportunity: a customer's incoming inspection team was trained to take a random sample of product and compare measured dimensions on the product with specified dimensions on the engineering drawing.
The inspection team went through a period where literally dozens of shipments from different suppliers of different products were put into quarantine because there appeared to be a discrepancy between actual product and drawing specifications.
Oddly, a material review board at the customer cleared all the quarantined product and attributed the wrongful quarantine as "operator error." In the meantime, of course, because of quarantines, product payment was delayed and the cascading effect disgruntled the customer's entire supply chain, especially since none of the products were ultimately found discrepant by the customer's own MRB. There were dark hints the customer was conspiring to stretch out its payments by use of a "suspected" discrepancy. The suppliers were so disgruntled, they formed a committee to face down the customer's top management over the issue.
Three supplier quality managers watched over the shoulders of the customer's incoming inspection team for less than five minutes before one recognized the factor causing all the erroneous data entries:
Other than an engineering drawing, the inspectors had no systematic plan for inspecting any of the products. No two inspectors measured the same part's dimension in the same order. There was frequent switching back and forth between inspection instruments (from micrometer to vernier to hole gage to thread gage and back again.) On some pieces an inspector might pick up the same instrument five times to measure different dimension on the same product. Worst of all, inspectors started with a blank sheet of paper and wrote in the specified dimension and tolerance for each dimension checked and then entered their own reading.
Looking at the accumulated data sheets,it could be seen that inspectors had caught themselves making incorrect entries, diligently put a line through the erroneous entry, added the correct one, and initialed the correction.
The error proofing solution:
each engineering drawing for a product had dimension features individually numbered in the order in which each was to be inspected, so that an inspector would make all vernier caliper readings first, then micrometer readings, then hole gage, then thread gage readings. Each product had a data entry sheet form with the specification and tolerance pre-printed along with the type of instrument to be used for the measurement. The inspector would enter his reading plus there was a space to record the ID number of the actual instrument used by the inspector (for tracking and recall in case the instrument was later learned to be out of calibration.) The system was duplicated at each supplier so everyone was literally on the same page when making an inspection of a product.
Net result:
an inspection process which eliminated nearly 90% of data entry errors and cut the net inspection time on a product by nearly 70%. There were no further quarantines of any suspect products later found to be OK.