Next Generation Product Development Tools, Part 12

I previously offered (reLink) that there were three issues of particular relevance to the future evolution of “next generation” 3D CAD applications:

  • Limited data portability
  • Proprietary, closed source formats
  • Insufficient support for extensible semantic information
  • My last entry, in which I discussed the emergence of “direct modeling” 3D CAD, was still very much about data portability. While I’d previously confessed to “largely ignoring the trend toward ‘direct modeling’ “, the truth is that in researching next generation solutions I found I couldn’t ignore the trend. And this is because the direct modeling segment of the 3D CAD market is solving the kinds of data portability issues which is frustrating so many users.

    There appear to be two things in particular contributing to the collaborative capabilities of direct modeling applications:

    1) Direct modeling files, by virtue of not having a parametric feature history, are effectively “dumber” than those 3D CAD file formats which do have them. Thus such files are more easily shared between applications using readily-available standard interchange formats: STEP, JT and IGES. After all, you can’t lose what you never had.

    2) Feature Recognition Technology (FRT) allows direct modeling applications to extract information inherent in the geometry (see Addendum below). And since FRT is, from my limited research, often based on a method of classifying faces, direct modeling methodologies seem to be a good fit.

    So what we have are applications which are not only more intuitive for many people to learn because they break from the parametric feature history paradigm, but we also have well-established interchange formats for them and techniques for extracting usable information from that data.

    One could assert solutions utilizing parametric history are more powerful, but if the data isn’t useful beyond a company’s firewalls or outside a small circle of vendors, their additional capability is arguably negated. I imagine this is the very reason direct modelers have increased in popularity. For many companies the trade-off is worth it, especially since most of these applications are very competitively priced.

    Remember the SolidWorks example previously mentioned (reLink)?

    As we saw with SolidWorks some time ago, a productivity enhancing user interface combined with competitive pricing can breach insular walls and drive “volume” migration/adoption. The consequence of course being that competitors are forced to respond in kind.

    I suspect this is what’s been happening over the past few years; pushing highend companies to respond by adding direct modeling features to their applications.

    Next up, I’ll try to post some examples showing what this push toward collaboration has yielded. Some of you might be as pleasantly surprised as I was.

    Addendum: Information regarding the relevant origins of Feature Recognition Technology was difficult to locate, so I’m including here a brief excerpt from the book Parametric and Feature-based CAD/CAM: Concepts, Techniques and Applications by Shah and Mäntylä.

    The initial motivation for using features seems to have emerged from a desire to integrate computer-aided process planning (CAPP) with computer-aided design (CAD). Research on techniques for providing data on manufacturing features for NC programming dates bacak to the mid-1970’s. Several researchers at Cambridge University’s CAD Center worked on automatic feature recognition from 1975 to 1980. Grayer, in his Ph.D. dissertation in 1976, presented methods for automaticing NC programming for milling (Grayer 1976). Grayer’s sectioning technique was applied to 2 ½ D milling, but it did not actually recognize features. Kyprianou appears to have been the first to introduce the idea of feature recognition by discovering topological and geometric patterns in CAD databases and comparing them to the characterizing features that needed to be found. The key idea was topological entity classification based on geometry (Kyprianou 1980). Almost all subsequent methods for feature recognition have used this idea in some form.

    4 thoughts on “Next Generation Product Development Tools, Part 12

    1. “And this is because the direct modeling segment of the 3D CAD market is solving the kinds of data portability issues which is frustrating so many users.”

      I should have read the entire series before posting my comments 3 posts back. This is pretty much the conclusion I was coming to.

      You didn’t mention Unigraphics in your previous post. UG had direct-modeling tools very early in its life, with parametric capabilities added on later. The result is a hybrid parametric/non-parametric toolset that is incredibly flexible but, to your point of the MCAD software development plateau, the interface is super inconsistent and, at times, confusing.

    2. Unigraphics = Siemens = NX5

      Same is true of CATIA. They’ve had direct modeling capabilities for a while. Both however have improved this functionality in recent releases, from what I recall reading.

    3. Ah, I didn’t realize that UG was part of the Siemens corporate family. I’m anxiously awaiting NX6, actually, so I can run it on my MacBook Pro. Nothing like taking work home, huh?

      I used Catia back in ’99. I think it was V5r13-ish. I loved it. I would have loved it even more if I had known about the direct modeling tools. I was taught to build models parametrically.

      Can’t wait to see where you go with the rest of the series.

    4. Pingback: reBang weblog

    Comments are closed.