Here is another example contrasting Watts and Garwood:
Watts - A company makes an electrical panel (Doc. No 123456, Rev. C, married to P/N 123456-03) that gets a circuit breaker in a rectangular hole. The breaker is discontinued by the supplier. The replacement breaker takes a D-shaped hole. Using detail views or notes or whatever method works best, the document is revised to show two configurations: It becomes Doc. 123456, Rev D and is now married to P/Ns 123456-03 and 123456-04. One way to do it on the drawing would be to have Detail A show the rectangular hole and note that it represents P/N 123456-03 and have Detail B show the D-shaped hole and note that it represents P/N 123456-04. So with one document, we can maintain a design history to support service requirements, especially if the implementation included accurate recordkeeping. Even companies that use tabulations (suffixes) often fail to change the part number. (123456-03 stays at 123456-03.) They just bump the document revision, as if a document revision controls interchangeability. (A very poor practice.) In this case, traceability is lost and the service department is embarrassed to ask the customer if they have the rectangular breaker or the round one. Service folks should know which breaker the customer has as soon as the customer tells them the serial number.
Garwood - A company makes an electrical panel (Doc. No 12345, Rev. C, married to P/N 234234, which appears nowhere on the document) that gets a circuit breaker in a rectangular hole. The breaker is discontinued by the supplier. The replacement breaker takes a D-shaped hole. Using detail views or notes or whatever method works best, the document is revised to show two configurations: It becomes Doc. 12345, Rev D and is now married to P/Ns 234234 and 414141. One way to do it on the drawing would be to have Detail A show the rectangular hole and designate it 12345-01 and have Detail B show the D-shaped hole and designate it 12345-02. But this method requires that recordkeeping (database relationships) includes a link from the document designation (suffix included) to the part number, including the dates where this match is 'effective'. (The document revision might give a clue to help establish traceability, but that would not be a proper way to handle it.) So with one document, we can maintain a design history to support service requirements, but only if the implementation included accurate recordkeeping.
I'll guess that most companies that have their act together use the Watts approach properly. But don't give up on the Garwood approach until you understand it well, because there is the opportunity for lean bills of materials and other efficiencies that can really please folks downstream of the design group. Operations folk will probably vote for Watts (because it is a mature and well-known methodology) when it is entirely possible that Garwood would serve them better.
Peter Truitt
Minnesota