In design automation circles, interoperability is said to be like the weather; everybody talks about it, but nobody does anything about it. Recent years have seen dramatic improvements in engineering data interoperability, but the problems persist. This report, from Engineering Automation Report in 2007, gives the vendors of interoperability solutions a chance to address the issues.

By Rachael Dalton-Taggart

Engineering Automation Report, May 2007—When I asked a reader this week if he has problems exchanging files between CAD systems, his typed response was “YES” in capital letters. The thing is, whomever you ask, no one says “No” to that question. Interoperability of CAD data, especially 3D data, remains one of the biggest unanswered and open-ended questions in the industry: “When will CAD data interoperate”? or “Why won’t the CAD vendors make their file formats open and interoperable?” have been asked again and again at events like COFES (the Congress On the Future of Engineering Software.) And regardless of the brilliant minds in attendance, those questions have never been fully answered.

So to try and understand, I posed these questions to the interoperability vendors themselves—people who have delivered successful solutions to this problem, and in the meantime carved out a market that in total is, we think, worth about $300 million to $500 million in annual revenues. Please note that they are representative of the interoperability vendors? market so their responses should be considered potentially biased.

When will CAD data interoperate?

According to the interoperability vendors, the answer can be paraphrased as “it already does” (via their products) but they seem to agree that if the problem were to be left to the CAD vendors, the answer would be “never!” Some see it as a prolonged technological issue. Others attribute it to the need for CAD vendors to protect their turf. Either way, the CAD interoperability vendors have a huge range of interoperability products and services out there that do work—most of the time.

“As long as there is more than one CAD system, there will be interest in translation,” says Peter Heath, product marketing manager, Elysium Inc. “The mathematical basis and maturity of the products still present differences that require detailed expertise to resolve.”

Phil Spreier, director of channel development at Dassault Systemes Spatial Corp, concurs. “3D interoperability becomes more complex each year. CAD vendors are driven to constantly improve and update their products. This necessitates the changing of their file formats…. Interoperability will not be solved until the industry either consolidates down to one CAD vendor or one file format (or at least no new ones.)”

“There are still common failures with geometry translation such as edges not stitching together, or normals pointing in the wrong direction,” says Peter Dickin, Delcam Marketing Manager. “However, these problems tend to be nuisance issues rather than serious errors and are easily repaired when they occur.”

Todd Reade, CEO of TransMagic, cites encryption, updating CAD versions, and also the rush towards new 3D Publishing formats as current issues that keep them occupied and which present continual barriers for end users. But he also mentions the CAD vendors themselves: “Few CAD vendors are interested in solving these CAD interoperability issues, and typically they don’t have the expertise to do it.”

But is it a form of protectionism by the CAD vendors? Chris Williams, CEO of Seemage Inc., a developer of 3D Publishing technologies, thinks so. “The CAD guys need to protect their franchise,” he says. “They simply can’t allow this [interoperability].”

How well does CAD interoperability work?

Go to any interoperability vendor web site and you will see claims of months of time being saved through the use of interoperability software. And I pretty much believe them. I have talked several times with the ING Renault F1 Team (a customer of Elysium) who have explained that they use a network of 50 suppliers on most of the 5,000 parts on a Formula 1 engine. By using CAD translation software, it saves them 2-3 days per part in turnaround time alone. By any estimate, that’s a hell of a savings.

“If interoperability software is good enough for airplane engines, then it should certainly be OK for toasters and printers,” comments Reade of TransMagic. “The 3D CAD market has matured in the last ten years. There are now several million new MCAD users creating 3D designs. The related 3D interoperability problems that were surfacing 10 years ago have only recently been addressed with reliable solutions… the technology is getting very solid and the advances are remarkable.”

The essential message we get from the interoperability vendors is that geometric data exchange (dare I say ‘dumb’ geometry?) works very well when using decent translation tools, saving time, effort and frustration. But anomalies remain. TransMagic’s Reade explains this well. “The two biggest interoperability challenges we see are IGES and Catia V4 files. And if you have an IGES file from Catia V4, that’s probably the worst-case scenario for data exchange. You better pack several lunches if you’re doing that job manually. Even getting data from Catia V4 to Catia V5 can be challenging if you don’t have the right tools. But if you do, the time savings can have an enormous (positive) impact on productivity.”

It is our understanding that geometric data can be exchanged easily, and in mostly usable forms, if you have the right tools. Quite often simply using IGES or STEP between CAD systems can be enough for geometric exchange, and every CAD vendor has IGES and STEP now, right? But to add complexity to this discussion, these same standard exchange formats can be affected by the quality of the translator inside the CAD system.

“The problem is not so much the CAD system as it is the data format chosen for translation and the quality of the data translator itself.” Says Spreier of Spatial, a company which provides CAD translation components for inclusion in CAD systems.

Reade of TransMagic comments, “Typically, the CAD vendors do not have the expertise to develop high quality CAD interoperability tools. Too often, they only care about adding a Check Mark on their spec sheet and thus, produce poor quality translators.”

“The glaring problem we see in data translation is ensuring the integrity of a translated model,” continues Spreier of Spatial. “The success of any data translation is not whether you simply get a translated model but that the model can be used without failure for an intended engineering process.”

If geometric data exchange is pretty much solved, what’s next in interoperability?

By all accounts geometric data interoperability is solved—and solvable—due to some pretty high-end development by the interoperability people. But if that’s “dumb” geometry, when will we be able to exchange “intelligent” geometry? Not only that, what about attribute data, Bills of Materials data and so on? Further, do users want to translate more?

“Is this a problem worth fixing?” counters Chris Williams, Seemage, when the question is posed. “People seem to develop products using multiple CAD systems and get past the fact that they cannot share feature-level information. This makes total sense since design is typically segregated to groups, who already have defined interfaces.”

Good question. Do the users need feature-level exchange of data? Some interop vendors are already tackling the problem. Proficiency was first to market some years ago with its Collaboration Gateway. (The company has just announced $5 million in new venture funding to help it make a comeback into the market.) Elysium has had its CADFeature product on the market for about two years. But all seem to agree that the issue will take more time before it is declared to be solved.

“The main areas for development are around providing a more complete representation of the original data in the target system,” says Peter Heath, Elysium. “This includes annotation, dimensions, drawings, colors, attributes, as well as the feature content. We always have interest from customers in a combination of products and services to allow anything from complete migration of data into a new CAD system, or exchange of more comprehensive data between parties.”

Peter Dickin at Delcam counters on some of this: “In many cases, having history and parametrics preserved can be a hindrance rather than a help when making minor changes to a CAD model to assist manufacturability. There can also be differences between design and manufacturing over how they perceive features. For example, the designer will see a network of reinforcement in an aerospace structure as a set of ribs, while the manufacturing engineer will see the same geometry as a set of pockets to be machined. Simple and even advanced translation cannot necessarily deal with those differing needs.”

But shouldn’t the CAD vendors take more responsibility for interoperability?

There is a general underlying feeling among users that the CAD vendors should make their file formats more interoperable. For example, the reader I was talking to earlier this week fights the realities of exchanging data between Catia and Pro/Engineer, as well as Mechanical Desktop and Pro/Engineer.

He said the following: “I have been after my PTC rep forever to try and get them to understand why [interoperability] is so painful, but it hasn’t helped much has it?” To further that comment, in December 2006, Autodesk and PTC announced an interoperability plan where Autodesk will incorporate PTC’s Granite kernel into AutoCAD and Autodesk Inventor, and PTC will incorporate RealDWG into its software.

The interoperability vendors, however, do not see the CAD vendors as providing an answer. In fact, when you look at it closely, maybe they don’t want the CAD vendors defining interoperability! Spreier of Spatial explains, “We don’t think that CAD vendor relationships are the answer because vendors have the ability to pick and choose what data can be translated to-and-from their file formats. These types of relationships seem a bit too much like leaving the fox in charge of the hen house. The advantage that third party specialists have is that they are driven to solve market needs, not maintain or capture market share for a CAD product.”

Reade of TransMagic adds, “You could say that there is a conflict of interest in the CAD industry when it comes to interoperability. Some vendors encrypt their file formats, creating new interoperability barriers. That’s why the problem exists. That’s why data cannot easily be shared with other systems. It’s unnatural to expect a CAD vendor to bridge the interoperability gap on a level playing field. It’s not in their interest competitively to be ‘open.’”

Williams of Seemage, adds, “Do you really believe these guys will trade feature-based definitions such that people can go back and forth between Pro/Engineer and Autodesk Inventor? I don’t!”

“Leaving interoperability to the primary CAD vendors would be like asking turkeys to organize a Thanksgiving dinner” states Dickin of Delcam. “The main reason that companies buy CAM software from the major CAD developers is the pressure from OEMs that have been convinced of the dire consequences of poor interoperability. Why would Dassault, UGS or PTC want to improve interoperability with a specialist program like Delcam’s PowerMILL? This would only reduce their CAM revenues! Most options that are available for using embedded software require the user to have a license of the host software as well. Thus revenues increase for the CAD vendors through the application of fear, uncertainty and doubt.”

An opposing point of view?

Now, having realized how biased these comments could be, I decided to ask at least one CAD vendor his opinion on this final question. Greg Milliken, CEO of Alibre, and always fun to work with, said the following, verbatim”

“We actually don’t think it is nearly as difficult as most CAD vendors would have customers believe. In fact, I think I agree with the interop providers that this issue is largely caused, or at least greatly exacerbated, by the CAD vendors. I don’t mean the effort is trivial, it is a complex process to exchange proprietary 3D model data between systems, but then again building these systems is hard too, so that isn’t an excuse.”

“If CAD vendors quit paying lip service to putting their customers first and made it a priority to make their file formats accessible, this problem would be dramatically reduced. Alibre put our money where our mouth is by publishing our format a while back with the STEP ISO group who controls such things, but of course other CAD vendors want nothing to do with actually improving interoperability. They want to lock in customers by trying to restrict their data.”

So now what?

The never-ending question about 3D CAD interoperability continues, but maybe at an advanced level to previous years. Geometry interoperability has been solved by some very bright people who have managed to work out the mathematics involved and incorporate that into some nifty solutions. Feature-level interoperability is here but requires some more advanced development to be appropriate for everyone. But here’s the dilemma: the interoperability vendors obviously have a financial interest in ensuring that CAD interoperability continues to be a solution people will pay money for; the CAD vendors have no financial motivation to provide across-the-board interoperability while they protect market share.

Ultimately, users are the only ones who can define what they want—either by making enough noise or by shunning products and solutions that do not provide what?s needed. But in the meantime, apathy, or a sullen acceptance of the current realities (and thereby finding work-arounds) is the reality of CAD interoperability.

At the time of this writing, Rachael Taggart-Dalton was Publisher of Engineering Automation Report. She is now Director of Marketing for Geomagic.