For decades, manufacturers have used "smart" or "intelligent" part numbers -- that is, descriptive part numbers that contain explicit indicators in the part-number string itself, indicating something about the name of the part itself, the product it is used in, physical characteristics of the part or other data points. One of the advantages of such descriptive part numbers has been that they allow employees to quickly identify a part and its purpose merely by reading the number.
As a report from Arena, a Foster City, Calif.-based developer of product lifecycle management (PLM) solutions, indicates, in the part number RES-100-0003 for a resistor, "100" could stand for resistance in ohms and "0003" could be a suffix indicating the part's position in a series.
However, some technology experts are warning that the use of such descriptive part numbers is not necessarily so "smart," and that they could drag down productivity in today's fast-changing manufacturing environments. A smarter tactic, they assert, is to employ auto-generated "insignificant" or "non-intelligent" part numbers and let information about the part reside in a database.
According to Dennis Gilhooley, senior consultant at Ultra Consultants, a specialist firm in enterprise resource planning (ERP) based in Chicago, descriptive part numbers' pre-computer-era origins might not be of as practical use today. "People would embed special coding into the part numbers," said Gilhooley, in an interview with ThomasNet News, "using a three-character prefix associated with the type of material or vendor or customer or something like that."
As an example, one of Ultra Consultants' clients is a large manufacturer of appliances that tries to embed multiple codes into its part numbers, including the company initials, the color, size, and other descriptors. What he has observed is that with many clients, "the numbers have continued to grow and grow" until they become hard to manage.
The long strings increase the likelihood of errors. "Between two parts, if there's only one number in a 15-digit code that's different, it can be easy to pick the wrong part," Gilhooley said. "If they're clearly different, that's a lot less likely to happen."
The best practice in part numbering today, Gilhooley wrote recently, is to assign each part an easier-to-manage sequence of five or six letters and numbers without any particular meaning. Since organizations manage their manufacturing operations with digital tools like ERP or PLM, insignificant part numbers "can easily be integrated into a parts database that can be accessed by materials handling, production, engineering, production control, purchasing or sales." In the database, any number of fields can be used to identify and describe a part with as much detail as needed.
This approach reduces training costs, as nobody has to learn how to decode part numbers. Knowledge about a part doesn't depend on the memory or experience of an individual employee, who might not be available when an identification is needed.
Another problem with descriptive numbering is that the description can become out of date and irrelevant over time. Individual parts can have their own life cycles; if a part has been identified according to the product, what happens if that product is discontinued but the part continues to be used in a newer product? Or what if a manufacturer changes vendors and the part number contains the name of the vendor that originally provided the piece?
Gilhooley admits that some Ultra Consultants clients have decided that switching from descriptive to auto-generated numbering would require too much organizational change. Some companies stick with old systems, and some opt for hybrid systems that perhaps retain descriptive numbers for existing parts but use auto-generated numbers for new parts.
On the other hand, PLM vendor Arena asserts that auto-generated numbering schemes can have their own drawbacks, particularly if a data entry error occurs. Suppose the person entering a part number into the ERP gets one character wrong? "With no frame of reference for a user to determine if the part number makes sense in the context of other data," according to the Arena paper, the error might be hard to spot.
Depending on the company's needs, there could be a compelling reason for employees to be able to quickly evaluate a part by looking at its number. For example, a descriptive part number might help a shop floor worker "to quickly verify that a part meets general functional requirements" or "to spot a part that's in the wrong group."
Steven Chalgren, vice president for product management and strategy at Arena, said that a wholesale migration from descriptive to auto-generated part numbers is simply not feasible for some companies. Speaking with ThomasNet News, Chalgren said that Arena's teams typically address part numbering during the implementation phase for their PLM systems. In the real world, manufacturers have distinct needs, and often companies end up going with hybrid systems.
Electronics manufacturer GoPro started out working with an original design manufacturer (ODM), who managed its part numbering. However, GoPro wanted to start managing part numbering internally, which it was able to do using Arena's PLM system.
Arena describes its part-numbering capabilities as "semi-automated" and "semi-intelligent," meaning it can support a hybrid numbering regime. In the case of GoPro, the Arena solution auto-generates part numbers but can also be tweaked to include descriptive information in the numbers. GoPro says the resultant solution allows for effective organization, filtering, and searching of part numbers.
"Since Arena PLM is almost always the originator of our customers' part numbers, we make sure we can handle many different part numbering preferences," Chalgren said. "While we have in our heads "~the best practice,' the reality is this is a gray area, and it really depends on the customer's products, industry, and preferences."