With all due respect, number of articles doesn’t imply quality of content.
Furthermore, I went to Advanced AUOM session at Insights expecting some complex UOM conversion and the example was a bike colors - no unit conversion. no planning sets, nada.
Respect Noted… I guess what I was trying to point out… There is a lot more content compared to when it was first released in 10.2.700, which I guess would not be hard…
I’m in an implementation right now, and boy do the decision-makers HATE the idea of sensible part numbers. I guess they allow 50 characters, why not use them?
We did half smart, half dumb part numbers. Users know/memorize the “smart” first 4 chars (which is basically a “product line”), and then a zone BAQ gives them a quick way to find and select the variant:
Our group resisted sensible part numbers for decades in our current system, but we’ve spent a great deal of effort standardizing this-go. Not least in the least to assist with 3rd-party CPQ & CAD integrations.
Luckily, we had a home-grown ‘part builder’ and they allowed a pair of us to go-for-it rather than try to make a million trade-off decisions as a large committee.
It’s a work in progress but I think they’ll like it. Expanding the core test group this week. We’ll see.
Make sure the people who designed the part system use it for a day. If people have to enter a part number 75-100 times a day, then make the designers type them in that many times in a row. Is the extra entry time worth it? Smart numbering system are designed for non-computer uses, usually reports or prints of some kind.
It takes no time to design a “smart” part number, but those part numbers will be used for an eternity - well, until the next implementation. I’m all for smart descriptions, but 19 characters seems a bit much to me, IMHO.
We have a varying range of part number lengths, however, the bulk of them are 12 characters or less. Our people would flog me in .25 seconds if they got any longer lol.
As one who does a lot of data research when things go sideways,I would not want to type those part numbers repeatedly.
While I do not advocate a totally random numbering system, I feel I need to point out that if you use all 26 letters, and all 10 numerical digits you are working in what can be called “base 36”… that means for each digit you get 36 values. The problem with base 36 is that you can spell naughty words (dont ask how I accidentally found this out with a 4-digit customer numbering system… oops), and so it is not really a good thing to use as a part numbering system that is exposed to the public… So… if we remove the vowels, this also removes the zero/letter-o problem… so that brings us to a base-31 system.
In a base 31 system, with 5 digits, you can acheive over 28.6 million values, which should be more than suffucient for any company instead of a 50 digit numerical value. I can probably remember 5 digits, but i cannot remember a 25 digit value.
Anyway, here is a base table that shows these counts:
Don’t let your users know about DMT. They will constantly ask you to do bulk data changes as part of daily operations. Sorry, I’m in IT…I shouldn’t be doing your job for you.
Has anyone actually ever done it? As in re-implemented the system they were already using, from scratch. When we moved physical location we set up a new site and were able to do a lot of cleaning up and improvements, that’s the nearest I’ve come to it.
We’ve seriously considered it for many of the reasons in this post, and a small part of me actually wants to try it, but most of me doesn’t have the courage to take it on. Plus getting buy in from the rest of the company who thinks it already ‘works ok’ would probably be difficult.