Citizens of CADtopia

CADtopia_JodiCADtopia must be a really nice place. All that endless storage, all those panoramic holographic displays, and not even a passing thought about CAD interoperability. Jodie Foster is probably on staff to ensure that all geometry and product structure information is passed along via a ubiquitous standard format. She remains ever watchful of unauthorized IGES use, or perhaps illegals translating data from one proprietary format to another. Violations result in revocation of citizenship, or in the case of illegals, the issue is efficiently resolved with armed robots, irritable mercenaries, or space missiles. Everything runs smoothly in CADtopia, and no one wastes any time re-examining translation chains or healing geometry. Then suddenly we wake up… Only to face the bitter reality: the persistent ridiculousness that is CAD interoperability today. CADtopia is but a dream.

CAD interoperability woes are rooted in the underlying math needed to bridge proprietary formats. The same formats closely coupled with the disparate internal models of each CAD kernel. Today’s market is defined by a multi-CAD reality: the variety and intermingling of the various CAD tools has never been more far-reaching, and there’s no turning back. Much of the pain is documented in Kenneth Wong’s great Desktop Engineering article The CAD Interoperability Monster Still Lurks. The CAD industry hasn’t solved the underlying interoperability problem, instead we manage it like some sort of incurable disease.

In the best cases, designs are exported into neutral and/or lightweight formats like STEP, PDF, or JT for sharing, reworked in the original proprietary formats and republished as necessary. In others, the new wave of direct editing tools are used to accomplish changes in geometry authored by another CAD source, but in many cases at the expense of design intent. In the worst cases, the models are just reworked. As a result of all of these cases, many workflows are necessarily one way, in order to avoid round-trip surprises, and the resulting math-induced headaches. Truly resolving the interoperability problem is unattainable, they say. It’s a pipe dream. We’re doing the best we can managing it with direct editing technologies and carefully validated quality processes. The price of multi-CAD freedom is eternal vigilance.

Maybe they’re right. But what holds us back from CADtopia is not some impossibly complex technical conundrum, it’s us. Or rather, the market. The market must decide to change or otherwise be coerced.

But who will be the catalyst? The hope of some independent standard being cooked up in an absurdly long specification is questionable, though many hopes are pinned on exactly that: STEP 242 ed2. It’s based on a 2009 initiative called <deep breath> “Development of a Convergent Modular STEP Application Protocol Based on AP 203 and AP 214: STEP AP 242 – Managed Model Based 3D Engineering.” Yes that sounds about as agile as it is. It would be nice if everyone could wait the thousand years or so for this noble attempt to come to fruition. The bitter reality is that such a classical standards approach will be ineffective. By the time the standard rolls around and gets properly implemented, it’s almost guaranteed to be obsolete. Assuming it’s even usable. One centralized entity (even if it’s a cross-industry collaboration) tackling the totality of the problem can’t keep up with every CAD vendor going in whatever direction they want to go. The pace of innovation is simply too fast and too broad.

If the standard approach won’t work what about the vendors themselves? As mentioned in the Desktop engineering article:

“Many software vendors recognize they should make a concerted effort to help users cope with the headaches of a multi-CAD environment. After all, their own collective welfare is inseparably linked to it.  As Siemens’ Paul Brown concludes, competing CAD vendors have to coexist: “It’s not in anyone’s interest for any one of us to make it really difficult for our customers to move data around.”

What a wonderfully optimistic thought. Gives me this sort of soothing yet uneasy feeling. Sort of like Jodie Foster singing a sweet lullaby to you right before one of the armed robots stabs you in the brain. From the same article:

“Getting all vendors to migrate to a common data model — the same formula for geometry creation and editing — is highly unlikely. After all, they’re fierce competitors. But there’s something else that gets in the way. The differences in their modeling formulas also give them the ability to develop unique features that they can claim to be exclusively theirs.”

Disparate models can be bridged. Let’s cut straight to the elephant in the room: intellectual property. This is about patent protection of the very technologies that define the livelihoods of the CAD vendors. Vendors will not do this willingly because it undermines their competitive advantage. They are fiercely competitive for all the right reasons.

Vendors are motivated to make their incoming translators as robust as possible. That’s where all the development focus goes with respect to multi-CAD. Focused on bringing the data in, with hopefully minimal problems, as long as you don’t send it out on a roundtrip. Round-tripping requires vendors on either end to effectively show their hole cards, essentially bridging alternatives to their own IP. The resulting consequence is an endless war of attrition, engineered towards a strictly hypothetical future where all customers are acquired by a single vendor and the problem is solved via monopoly. You could call that the Microsoft Office scenario. It ignores the central issue: the multi-CAD universe is here to stay.

So it’s safe to say you can’t count on the vendors to resolve the issue on their own accord, unless perhaps if you dissolve all software patents (as some suggest). That’s an approach I consider to be heavy-handed; i.e. the nuclear option. Vendors could also voluntarily open source, but again what would be their motivation? Unabashed belief in CADtopia?

A third scenario could be a mandate, and I’ll talk about that possible scenario next time, when I bring digital stewardship into the mix. In the meantime tell be about your dreams of CADtopia, and whether you think it’s attainable.

  • Joe Brouwer

    I have been beating this drum for over 20 years. I have worked with solid models from all of the popular systems. You can easily read a CUBE in any system without any problems, but once you step out of just simple extruded and revolved parts or even add a hole, you are into the limitation of the CAD system in exporting the correct part.

    As all of you “Users” of history based Pro/e and its clones know that the skill of the user is everything. A history based only part can be a huge problem to edit, it is tough for the new or marginal user to edit a complex expert created part. Many parts have put the designer into a corner, making them use some very bizarre processes to solve the problem, causing you experts to scream and pull your hair at an amateur created part, of course, we are all experts. LOL.


    I have had SW and Pro/e parts and I have given up on in the native system and just exported them to a direct editing solution. I am not an expert user in either of those programs and I found many parts impossible to understand. I tossed both programs aside years ago. My tool of choice is IronCAD.

    Universal CAD Compatibility is Here!!.html

    I had a Boeing Catia file with little .010 ridges all over the part. You could see where the designer just gave up. I had to help the CNC fellow getting this model machinable. It was a very corrupted part and we did it by cleaning up small areas at a time, since we could not fix the complete part. I wonder how they inspected it.

    I have had Pro/e and SW parts that were corrupt and had to be corrected in direct edit programs. Sometimes you need a hybrid modeler to make this happen. I have had experience with Catia 4/5, Pro/e, NX, Inventor, SE etc. Lucky I have had Hybrid modeling capabilities available, hybrid modeling was required many times getting down to wireframe to correct the parts. Then healing the surfaced model into a workable solid. Actually SW has great healing capabilities.

    Take a look at the first example in this article. I was designed in Pro/e, truly we had no problems with the imported file. This job was done in 2004, yes, 10 years ago. You all act like this is a new situation.


    Beyond Direct Edit: Surface and Hybrid Modeling

    It would be interesting to see how the direct edit only packages hold up. Maybe take the direct edited files from Creo Direct, IronCAD, NX ST, SW Mechanical Conceptual and Spaceclaim and see how compatible the parts would be. With a variety of formats. I have a nice couple of drawings of some parts to create for the test.

    Now SW, Inventor and ZW3D have direct edit modules. It would be fun to see if you can design in them and export the files. I realize that the SW users are basically unaware of this functionality and are not using it in their design process. But it is quite powerful and fun and easy to use.

    Are Solidworks Users Stupid?

    If any of you have seen SpaceClaim in action. It is truly an incredible package. It was the only package designed from the ground up as a direct editing package. I tried to sell it but it basically duplicated my existing products at the time. I have never warmed up to a direct edit only package for conceptual design. Here is a job I did with SpaceClaim.


    I feel history based design gives more control over the conceptual design process. I get in trouble because I sell IronCAD. I sell it because it is the best conceptual design package on the market.

    CONCEPTUAL DESIGN – Which CAD Paradigm is Best?

    It is not that the CAD vendors don’t want a common compatible format. It really does not enter their priorities. Most small manufacturing companies only used one CAD system and rarely have an OEM parts and if they do, all of today’s CAD systems, besides Catia 5 provide tools to directly edit if necessary, but it rarely is necessary. But the CAD vendors seem to be running scared. PTC seems to be the one that is worried most about it. They don’t have a mid-ranged easy to use system to fall back on and seem to be worried about keeping their users. See my articles below about the Best to Worst CAD systems.

    Why would I know this? I have sold, supported, trained PC based CAD and provided engineering services since 1986. Over those years I have supported most of the Boeing suppliers, first with Catia 4 then the horror show of MBE and PLM with Catia 5. And I still do. I am not some self-claimed expert pundit watching from some ivory tower touting his/her opinions. I have thousands and thousands of hours working with CAD in a variety of systems starting with wireframe. Today, I use 3 systems interactively. None are the clunky Pro/e Clones.


    The Boeing’s and Airbus’s of the world using Catia 5 are having a horror show not being compatible with the industry. Catia 5 is the worse program for airplane design as a basic CAD system. It is really shocking to me that Dassault did not release the SW Mechanical Conceptual as an add-on for Catia 5. It is based on the Catia 5/6 modeling kernel. I really don’t understand Dassault, but then what CAD vendor do I understand???

    The Worst to Best CAD System and Why!!%20-1.html

    I can tell you that the users in many large companies are totally ignorant of the interoperability needs of the industry. They are constantly sending out IGES files without any consideration of what the suppler needs. This is a constant problem with my Boeing supplier customers. The supplier is the only that really needs a standard solid model format. You have to get that model duplicated into the CNC software.

    Lost In Translation! A Guide to CAD Translation Formats.

    But it is not just to provide the supplier with a compatible file. It is to open the industry up to a standard process. The industry in the past was based on a proven standard. It is shocking to me that to computerize this simple process is in such a quandary. Not only the problem with the data but the different CAD systems have made expertise in the system a priority over industry experience. Only in single part design has CAD provided a cost effective solution. The rest from assembly to data management has been a bastion of vested interest costing the industry Billions of $$.

    You all may enjoy my latest article.

    The Death of the Drawing

    • I’m going to call the folks @ Guinness, that may be the longest comment known to mankind.

      • Joe Brouwer

        I am not sure to say thanks or…..

        It is a topic I have been involved in since 1985 after being introduced to CADKEY at Boeing. It was the first time I trading “3D IGES” wireframe files between CAD systems. We easily moved files back and forth between CADKEY and Catia. Then I went to Square D and did the same with Computervision CADDS 4. Selling them 3 seats of CADKEY as a VAR.

        This was pre-dxf!!

        Hmmm Dxf??… Nah.. give these people a break!!! I feel a new article coming on!!

        You know, I never realized how historic that was!! I have been working with CAD data transfer for almost 30 years. How old are you ED?? LOL.

  • Pingback: Escape to CADtopia | E(E)()

  • Pingback: Lost in CAD Translation | E(E)()