Throwing more gasoline on the part number debate

There's always a little room to throw a bit more gasoline into the fiery part number debate, the ongoing saga between usable human interfaces and computational perfection (coming to a theater near you this summer). After all, delightful never-ending arguments over part number intelligence don't have to be limited to just part numbers. Turns out we can argue and/or rage about part descriptions too. Are part descriptions a critical delineation amongst a sea of look-alike parts? Or are descriptions otherwise considered to be interesting, but irrelevant, acting only as a hopelessly outdated substitute for classification metadata? Let's explore.

SONY DSC

But what is a part description, anyway?

That's actually a more complicated question than you might think. In theory, a part description should be just like what it sounds: text that generally describes the nature of a part, but unlike a part number is not necessarily unique. A description helps you understand more about a part without having to look at that part's engineering in detail. A bit confusingly, most engineers, knowingly or not, conform part descriptions to a schema defined by a variety of sources, including current ASME standards that were originally created for drawing title. The familiar noun and noun phrase with modifier pattern is derived from the original mil standards, crafted from an age when file cabinets ruled the earth.

Adding further to the confusion, some refer to the description as part name, and depending on the CAD and/or management software in use, part name, description, and drawing title may be the same thing or might be entirely independent fields. So are we talking about part name, part description or drawing title? In a word, yes. But your mileage and practice will vary. Regardless, more often than not part name/descriptions/titles are freeform text fields. And whenever there's freeform text available, of course wackiness ensues.

So that's just peachy, but do we really need part descriptions? Depends on who you ask.

Down with part descriptions

The system architects will argue that humans left to freely romp in the uninhibited wilderness of free text fields, invariably undermine the utility of a part description. Typos, outright errors, and the occasional improvisation undermine search performance and effectiveness, which is kind of the point of having a part description to begin with. If this rationale sounds familiar, that's because it's the very same argument used to snub intelligent part numbers. Ironically, many who want to get away from the downsides of intelligent part numbers instead load the part description with intelligence.

That merely shifts the problem. Oops.

Description haters would argue that a conventional description is nothing more than a poor man's classification, a "good enough at the time" solution before computers brought us robust classification taxonomies. In fact the folks at ZeroWait-State take it a step further, insisting that a robust classification system could be used to automate part description entirely, ensuring that content is both standardized and vetted, but still remain easily human-readable.

Part descriptions forever

But others would argue that free descriptions should remain so, that engineers need a "combo breaker" on top of any given classification methodology that can differentiate parts that end up with the same classification. Like that one time you sorted through 70 septillion aluminum flat washers. While the classification taxonomy could always be expanded to include the necessary category differentiators, it's not always easy nor accessible for the engineer to do so, and if that taxonomy becomes too deep and complicated, there are computational and usability prices to pay. Categorization strategies breakdown when there are a large number of uniquely differentiated parts, that vary from others in dimensions not yet categorized and that will not apply often in the future. Open-ended descriptions allow flexibility to load all that information in the description in whatever way seems most efficient.

Ultimately there's not one-size-fits-all solution here, the types of products you create and especially what you do with your part numbers will factor heavily in how you use part description. Whatever your own opinion might be, it looks like you might have to take sides once again. We'll be waiting for your comments in this durable-looking bunker made entirely out of aluminum washers.

 


 

  • Ryan

    Ed-
    Great posting, as usual. You won’t be getting any incoming phone calls to help my drone locate your position and drop washers on top of you from 1,000 ft. I agree that one size won’t fit all, if even three! You just need to have a serious discussion with your organization as to how each field (part number, description, drawing title..not to mention your CAD attributes) will work within your current/future systems (ERP, CAD attributes, PLM metadata, etc.). Once this is understood then go forth and populate and keep everyone on the same page.

    • I for one am most glad that errant drones dealing small, thin, and circular death from above have not been dispatched.
      You’re right about ensuring an organization has a good internal discussion to find the right balance given their specific technology landscape. But sometimes potential requirements get glossed over – how often do you think this doesn’t happen?

      • Smaller manufacturers struggle with this a lot. With 1-3 engineers, it’s so easy to just let them do it “their way” instead of reigning it into a consistent system. Especially considering many of these smaller places lack an ERP system.

        May washer 92150A101 come raining down on you. At least it’s light :).

  • I come from a manufacturing background parts are generally manufactured one time and then never seen again. I have found that the part number is important, but a good description and tags are good things, too. The part number is handy for communicating EXACTLY what’s being made. The biggest challenge I’ve had is finding a design that was made 3 + years ago to be re-used for one reason or another, and I’ve found that a good part description, and tags have been instrumental in helping me find parts deep in the archives.