Good point, Nate. I think David gets at that in his post, as you well noted. The bottom line is the standards bodies can make capabilities available to the community at large, but it really comes down to whether the vendors take advantage of it. Back to David's point--it's the unfortunate reality that many times they don't in order to protect what they see as their best interest and their intellectual property.
After reading all of the replies, I tend to agree with David's most recent summary. The vendors are coming around, but slowly. One of the issues not mentioned directly in any of the posts (unless I missed it - apologies) is the discrepancy between what the standards and specifications allow for and what is actually implemented by the technology provider. Semantic intelligence, geometric constraints, product structure, etc. are provided for in current or upcoming releases of STEP (and I think JT), but it is a question of whether or not the technology provider implements that functionality in the deployment of their translator.
There are standards bodies and CAD interoperability standards out there and have been for years (IGES and STEP being the main ones), but they never seem to fully address the compatibility and model sharing needs as well as users would like (although David Prawel's Interoperability survey shows that use of these standards is on the rise). It's likely the same as with any standards body--they move more slowly and more often than not, can't serve everyone's needs.
Are there any standards bodies which can play a role in fostering interoperable CAD data file standards, or is it more about the vendors themselves responding to user pressure (as well as realizing that, if it's in their customers' best interests, ultimately it's in their interest, too)?
Thanks for your thoughtful response, David. I think your point about there being no need to send full CAD files all around the world for collaboration is important and one the vendors seem to be latching on to to promote this idea of multi-CAD interoperability. PTC, in its new Creo release, played up this idea as part of its new functionality as did Dassault in its V62012 release. Lightweight and open file formats like Siemens' JTOpen also have a strong story. It may not be the complete solution, but it is progress.
I agree standards are the leading method to share data. I wish there were standards for creating not only data, but, rules, formula and knowledge. Today, a parametric model contains a variety of data, tables, rules, formulas. It is not just the geometry of parts, we are talking about. Today, many high-end PLM models are very rich. It contains all sorts of behavioral and engineering contents that is well beyond geometry. There exists no uniform standards for those and I do not know of any standard body that is looking into defining standards for exchanging such rules and knowledge across different CAD systems?
I agree today each vendor's solution is supported and embodied by their format. May be it is a good thing for that vendor. But that certainly puts unnecessary & often heavy burden on the users of such format (if they have to deal with multi-vendors), which is not a good thing.
It does come down to business priority, but it isn't really a disconnect. In defense of the CAD vendors, they have very expensive and very complex, internal code that defines the way their modelers do things. And to give others access to this is like, well, giving away your core asset. To use your analogy, someone devising a smart speech recognition system would not tell their competitors how they encode the speech, or interpret grammer - this is their crown jewels. The CAD guys have huge investments by hoards of PhDs to write tha interpret how models should be created and managed in their systems. You can't really ask them to give this up.
THis being said, it's also reasonable to expect them to play along with others to help users deal with their multi-CAD reality. To support the numbers Beth quoted in her article, our annual Collaboration & Interoperability Survey consistently shows that most companies work with 3 different CAD systems on a regular basis, and the majority deal with more than 3 different data formats, sometimes 8, to get their work done. Some of the CAD vendors have played along well, trying to help the community develop standard approaches to getting access to their internal data. I give credit to the Siemens guys for probably being the most active over the long haul in working with the standards bodies to give needed access, in a form users want it. In recent years PTC, to their credit, has also been a very visible and cooperative player, particularly in the STEP community. Notice I didn't mention Dassault, as they are only recently waking up to the fact that not all designers and engineers use CATIA - a topic for many more blogs perhaps...
Years ago the CAD vendors all took a fundementally closed view in this regard, but they have come to realize they win if they play an open game. So what's the answer you ask? THere is no single solution. It's true that standards like STEP are your best long term bet (including long-term archiving), but it's also true that it takes time to get standards bodies to move and adapt. STEP second edition has made a huge difference to many companies. BUt in the end, each user company must determine their specific requirements against what is possible in the standards community, and overlay that with their timeframe.
Case in point - there is no rason to send full CAD data files all over the world, when someone may only need a visual idea of the product/part. In many/most purchasing situations, for example, the RFP data packet doesn't need (and often cannot contain) CAD geometry, only a good quality picture that can be sliced and diced, sized and annotated. Editing the model itself is undesirable, usually not allowed, for good reason - who is responsible for the edits if the product fails? So technology like 3D PDF and JT fit this bill nicely. JT has made significant headway in the automotive space, and 3D PDF, well, suffice it to say, who doesn't know what PDF is and where 3D PDF might fit?
There will always be bizillions of data formats. Each vendor's solution is supported and embodied by their format. To them, it's the best way to handle their data structures. To others, it's not all that much value. So standards are the leading method to share data - again the most preferred exchange mechanism in our annual survey, especially among suppliers.
So ponder this one... do you own your data if it's in a proprietary format which is legally owned by a vendor?
All valid points and ones i've heard continuously. So you're saying the industry standard formats like IGES and STEP don't the distance in terms of solving the problem. What about some of the newer lightweight data sharing formats like Siemens JT or some of the hybrid direct modeling/parametric-based solutions? Any relief there?
I think disconnect lies in setting the business priority. I do not think Technology is the barrier here? When people can devise smart speech recognition system, when people can devise smart phones into a size of a cigarette box and when people can send man onto the moon, why it is so difficult to build a universal smart CAD model convertor?
As I said earlier, it all boils down to business priority. It is difficult to be done by an outside body though, since such bodies may not have access to the underlying API or GUI that they may require to build such a convertor. Even though they (outside bodies) can write such a working system (first) but to keep up with ongoing changes, the CAD vendors may possibly make in future, could become a daunting task (maintenance) for them.
I think how to entice our CAD developers to pay more attentions to building such smart CAD converter is the big question?
The vendors will say they are hearing users' frustrations loud and clear. Each major player that I talk to say they understand that having a single, monolithic CAD (and even PLM) system doesn't cut it in today's world of cross-company, cross-continent collaboration. So if they are indeed listening, I wonder were the disconnect still lies?
Are they robots or androids? We're not exactly sure. Each talking, gesturing Geminoid looks exactly like a real individual, starting with their creator, professor Hiroshi Ishiguro of Osaka University in Japan.
For industrial control applications, or even a simple assembly line, that machine can go almost 24/7 without a break. But what happens when the task is a little more complex? That’s where the “smart” machine would come in. The smart machine is one that has some simple (or complex in some cases) processing capability to be able to adapt to changing conditions. Such machines are suited for a host of applications, including automotive, aerospace, defense, medical, computers and electronics, telecommunications, consumer goods, and so on. This discussion will examine what’s possible with smart machines, and what tradeoffs need to be made to implement such a solution.