The One True Part Number System

The debate around part number systems is often reduced to a pitched battle of absolutes, pitting the cold simplicity of the Generic Numbering Coalition (GNC), against the historically focused Confederacy of Intelligent Numbers (CIN). There’s no shortage of arguments about one approach over the other. But you don’t have to play that game; there are always other possibilities. Let’s understand why.


The GNC was founded on the philosophy that part numbers are inconsequential. Part numbers are merely unique markers that point to a collection of metadata, and people shouldn’t bother with them. Let the machines do the work, keep humans and their fat-fingering ways out of the loop. After all, even the most elaborately designed intelligent part-numbering system is a temporary triumph, which will succumb to degradation, confusion, and irrelevancy over time.

CIN, on the other hand, upholds a humanistic appreciation of history. Part numbers are important; they will always be the subject of conversation, collaboration, and argument between humans and not systems. And to facilitate those interactions, part numbers must provide some system-independent context in of themselves because there comes a time we have to look up from our screens and communicate one on one. The only way to build such a context is to architect a recognizable system, and one that lasts.

Which System is Best?

Neither. Both approaches are myopic because they are extremes that wholly ignore the other’s chief advantage. A balance is needed. A hybrid methodology. But it’s much more than just crafting a semi-intelligent system and calling it a day. In order to understand precisely why, consider the following engineering realities:

  1. Defining a part number should never be an obstacle.  An idea should not be derailed merely because a person has to sit around and think about what a number should be or how to get it properly assigned.
  2. Flexibility is Key.  This is especially important for new design. Sometimes you don't know what you're building until you build it. Don't force decisions that don't have immediate answers. If your part numbering system has rigid classification you may find yourself stuck. The most resilient rules are not absolutes, but can bend when the situation requires.
  3. You should be able to remember several part numbers.  If you can't, then congratulations -you've built a system to welcome our new robot overlords, but not for people to use.
  4. Realize the system is not ubiquitous.  Relying on the system for certain information is necessary, but understand there will always be gaps in the system. We tend to oversell the utility of mobile tablets, phones, etc. in this regard. The reality is we’re going to talk about parts in dark corners devoid of technology.  In the hall, on a phone, under a plane, in an adhoc meeting sidebar, or in the vice president’s office. In none of these situations does everyone in the room have the luxury to sift through a device and sort some classification, just to tell two numbers apart. It's best to think of it almost from a military standpoint - what do you need when you're pinned and comms are down?
  5. Out of the Box (OOTB) tool behavior or current systems should not limit your solution. You need to fight the limitations to balance your objectives. But only if it makes sense.

But how do we incorporate all of this into something viable?  Here's a thought: how about creating an evolutionary system:

  • Use a temporary pool of simple numbers to remove obstacles when working fast and loose, especially in early-stage design.
  • Allow for easy re-identification (you'll have to fight the tool here).
  • Design gaps in numbering to permit grouping some numbers together when it makes sense to do so.
  • Let the numbers evolve as the design matures, and solidify those numbers into something that is meaningful through a release process.
  • Construct them with just enough information and structure for readability, keeping digits in groups of 5 or less.
  • Weigh each piece of intelligence carefully, and focus on properties that are both a) useful and b) immutable.
  • Consider three bits of information a maximum and not a goal.

So, what do you think?

If you can balance the above guidelines just right, you might have a winner on your hands. Of course, some of you may have your doubts. Perhaps you're a strong proponent of one side of the debate or the other. You probably have stories to tell, then. Feel free to share them in the comments.

guide to CAD file managementThe Next Generation of PDM  

More teams are using Cloud, Analytics, Mobile, and Social tools to speed up product development. Independent analyst firm, Consilia Vektor, explains how this changes Product Data Management (PDM) as you know it and how this can help your team work smarter.

guide to CAD file management