Part Numbers. Intelligent & Dumb. Can we have both?

Part Numbers. Intelligent & Dumb. Can we have both?


I think, most of PLM people are considering the fight about pros and cons of intelligent numbers is only related to product development and manufacturing. But the problem is more generic. It is about every number in enterprise software. Account number, order number, support ticket, etc. Thanks for Ed Lopategui that brought my attention to the following article – IT’s dumb fight over intelligent part numbers by David Taber. The article brings an extended perspective on why usage of easy memorable numbers can be beneficial. Read the article and draw your opinion.

According to David, new technology and cost trend can help us to afford to have both numbers. It will preserve internal identification to be referenced in databases, but generate significant and intuitive number for end users. Here is the passage that explains that:

Whether it’s a part number, an account number, an order number or an identifying number for nearly any real object, the users ask for a number that isn’t abstract, arbitrary and essentially meaningless. They ask for numbers that are short, significant and “intuitive” for the business user.

What has changed is the cost of storage and computation: redundancy and denormalization is essentially free, as long as it’s automatically maintained. Given that magnificent (and continuing) declining cost curve, we can now afford to have it both ways. We’ll keep with the impossibly long (and hard to remember) identifying numbers/strings that IT has already built, and those will be used as the foreign keys for all the tech. But now we can add a number designed just for humans, and which all areas of our tech are told to ignore.

I found the idea interesting and certainly applicable in some places. It made me think back about my yesterday article where I discussed the idea of part number transparency. Clearly, intelligent part number as proposed by David Taber’s article can provide a potential answer to the question what number should be displayed for better user experience.

The idea of having both numbers is fascinating. I can certainly see how PLM and ERP systems can “generate” useful Part Number based on some internal attributes, properties and other related information. At the same time, system will keep dumb number and use it for relationships in BoM, product configurations and ECOs.

At the same time, it made me concerned to think about the “temporary” nature of significant number as it proposed in the article. I guess the idea can work for account numbers. System can generate easy memorable account number that can be used in support, sales and related activities. What about Part Number that used for product configurations, bill of materials, component databases and many other places? How comfortable engineers will be with the fact each part number in BoM can be technically different in a year.

What is my conclusion? The low cost storage and computation in modern technologies can solve many problems of data management. Very often we ask for the wrong things and, as a result, getting wrong answers. It seems to me engineers that are asking for significant part numbers are trying to open the door that already open. It is up to the PLM, ERP and other systems to realize how to implement it in the way intelligent numbers will be displayed for users and strong references will be used to preserve data relationships and consistency. Just my thoughts…

Best, Oleg

Image courtesy of Stuart Miles at


Share This Post

  • This summarizes what I try to profess when the part number argument comes up.

    It doesn’t matter how good the technology is; at some point, somewhere, someone is going to have to consume that data. Until Boeing patents a fully automated assembly line, there will be humans in the loop consuming that data, and they will be too busy with drills and riveters to worry about smartphones and laptops. They will not have access to the technology and therefore the data has to be intuitive. Not necessarily smart, just intuitive.

  • beyondplm

    Scott, thanks for you comment! It does match exactly what I think – people want to have a consumable data. Not dumb numbers. And until we have people in the loop, we need to provide these numbers in a reasonable way. Best, Oleg

  • Oleg, I think you and Scott are on the same side. He’s stating that in production reality, the technology is not ubiquitous, so P/N that are difficult to work with barring access to the system can be troublesome. I’m sure if I’m paraphrasing incorrectly Scott will correct or alternatively throw a chair at me.

    The struggle here is that we’re taking what’s meant to be a database index and forcing everyone to work with the index. How many other databases do humans interact with the index? More thoughts coming soon… the new post should go out through GrabCAD next week. It’s about the end of part numbers.

  • beyondplm

    Ed, I think we are on the same side with Scott too. Maybe I didn’t explain myself well. P/N (or whatever else human readable) is required until humans are involved into the operation.

    Look forward to your post. “End of part numbers” sounds interesting 🙂

    Best, Oleg

  • Oleg, nice article. You are fully right regarding wrong questions. The question should be

    – give me the possibility to classify my product and to use this classification in all my processes

    and not

    – give me intelligent number.

    We implemented in some projects 2 numbers: main ID and alternate ID. The search function should look for both of them so that the product can be found. I have seen a nice example in a small standard part software, where you are able to switch flexible what you want to see: main or classifying ID.

    What I am not really clear about: what are the risks using such solutions?



  • beyondplm

    Alexander, thanks for your comment! There is no significant risks to have multiple numbers. The problem is not to have numbers, but to figure out how software is handling it. In general, the logic you implement (I guess) is pretty much custom. So, questions of uniqueness and support of selections based on these numbers are those I’d be trying to focus on first. Btw, what PLM product are you using? Is it TC?

  • yes, TC.

  • Pingback: Beyond PLM (Product Lifecycle Management) Blog » Will machine learning eliminate part numbers?()