The Nonstandard Standard

nutplateOne of the oldest Computer Aided Design (CAD) conundrums dating back to the invention of components (or was it electricity?)  is our close and dear friend, the so-called standard part.  Oh MS21075  nut plate, where art thou?  The days of a single designer rummaging around a dumpster (i.e. the mysterious misc directory) are arguably long gone, replaced with the relative opulence of curated standard part libraries.  But is the standard part library the final word on use and availability of commoditized part models?  It’s interesting that standard library concepts have remained largely unchanged in recent years, despite a completely new universe of developments including cloud and Big Data, technologies which already stand to revolutionize CAD data sharing entirely.  While fervent discussion about cloud virtualization or authoring in the cloud continue, very few want to talk about poor old MS21075.  No one loves him anymore.

The concept of standard parts is simple: prevent engineers and designers from literally reinventing the wheel.  Given that most any design has a certain percentage of commodity parts purchased from suppliers, i.e. fastener hardware, electrical components, what have you, there was tremendous value to be recognized in making sure models of that hardware were readily available to a CAD designer.  In the dark days of the late cretaceous (probably last month for some), CAD designers would either remodel or borrow whatever models they managed to scavenge.  Often these models were either incomplete or downright incorrect, resulting in costly design or manufacturing errors down the line.   As Product Data Management / Product Lifecycle Management (PDM/PLM) deployments were adopted, they provided a unique opportunity to rectify these long standing problems (with varying amounts of pain and suffering).  Despite being somewhat a side effect of the data cleansing necessary to transition native directories to managed environments (duplicate part numbers, anyone?) the case had value.  The end result was the emergence of curated standard part libraries.  Even in PDM-free environments, most of the modern CAD packages offer some sort of standard part library either straight out of the box or at additional cost.

Standard part libraries these days are common… perhaps too common.  They are everywhere, whether it’s the libraries built in the various tools, or one of an increasingly dizzying array of online providers or services which vary in cost, part availability, platform, format, and relative probability of how often you will be attacked by a vendor salesman.  Think that era is over?   Just a couple months back, the AIA and IHS, most famously known for charging everyone for access to  standards documents originally developed with public money more than half a century ago, has announced yet another player in the standard part model game.

So let’s think about this, instead of every designer out there wasting valuable time remodeling things that have been already remodeled we have successfully replaced that inefficient system with a variety of institutions and services which waste valuable time remodeling what’s already been modeled.  And then charge for it.  That’s progress.

Of course, that’s a bit harsh.  Use of standard parts libraries have been important in reducing design cycles and eliminating error counts, but it’s time to think bigger.  Especially if you’ve ever had to merge two libraries from company acquisitions; you’re probably thinking along the same lines.  CAD sharing in the cloud has provided an opportunity to make standard parts truly universal – there should be no middleman between those responsible for the part or standard and those that use them.  The trick is bridging the private clouds where the designs largely reside and the public clouds where the standards belong.  Not to mention all the format shifting (topic for a different blog).  Not an easy trick.  In fact, quite a hard trick.  I would hope the various CAD sharing cloud players would be right on it – but that hasn’t been the case.  GrabCAD’s for example, long ago explicitly stated they don’t want to be another standard parts library.

“Standard parts are useful but boring.  The overall trend is, that unlike the useful but boring standard parts libraries, GrabCAD is becoming two things – a lively community for engineers to share their work, inspire and help each other, and a global directory of engineers.”

Ouch.  I thought I was harsh.  Poor, poor MS21075, you are soooo boring.  Homer sleep now.

GrabCAD’s intentionality was necessary, if they had been labeled “just another standard parts repository” that would not have been good for them.  Regardless, the larger problem is still out there.  Case in point (Not to pick on GrabCAD because I actually quite admire them) but just the other day, they ironically talked about standard parts relative to Workbench:

“Many companies have their own parts libraries or templates that allow designers to use a single component in multiple designs without having to re-create it in each assembly file. This can save hours or days of work per project. Workbench can manage these common components, keeping them updated for all team members in all projects. The steps below show one example of how to implement a standard parts library in Workbench”

Interesting, that.  Because those parts are boring.  But what was that other word… useful?  Are we going to continue to replicate standard part repositories ad infinitum in every company server and across the cloud, or perhaps can we converge to the oneWhoa.

  • Thanks Ed, made me think of the fact that also PLM is boring . And strangely you become more profitable thanks to boring stuff

  • pgarrish

    I suspect there are fewer companies than you would expect (or hope) running decent standard parts libraries. I worked briefly with a company called Shapespace who have a geometric search tool. This can be used to collate the parts for a library (as well as finding duplicate drawn parts). Analysis at a couple of large companies discovered horrendous levels of duplication (20-30%) or near duplication and with companies re-structuring through mergers, separations, projects, system consolidation etc, I agree there is a huge opportunity not only to tidy up ‘standard’ parts but also drawn parts.

    As you point out, getting this stuff right can save a fortune – I doubt many engineering firms would question a cost of £500+ for designing a part – but if you add in the through-life costs of over-inventory (20 x 2 parts could probably be served by 25 x 1 common part), technical publication wasted effort, poor procurement (smaller batches) etc.. then there re literally £££M to be recovered….. boring indeed!

    • The geometric search tools can be very powerful and you are right, the amount of duplication I have witnessed out in the field is staggering. Enough reinvented wheels to start a wheel festival.