Anticipating the Next Paradigm Shift in CAD

By Kenneth Wong

Nearly a decade ago, feeling claustrophobic in a stuffy conference room, I had my first encounter with parametric CAD. I was a new hire for Cadence magazine (now extinct), a greenhorn surrounded by seasoned editors. As I watched, a product manager with a laptop proceeded to shape a cylindrical tube into an automobile muffler. He created an outer chamber, resized the pipe inside, punched holes, bent the mounts, and then spun the finished model for us to admire. Throughout, he emphasized the software’s intuitiveness.

To me the workflow was anything but intuitive. I had questions: How would I know where to find the extrusion command on that cluttered menu? Why do I need to go back to the feature tree just to move the pipe a few centimeters? But I seemed to be the only one not grasping the software. My learned colleagues appeared to be in agreement on the software’s ease of use.

Eventually, parametric modeling did begin to feel natural to me. I became accustomed to its logic, so the required steps—select a feature, pick a command, define the direction of the edit along an axis, enter a numeric value, and so on—made sense.

Years later, I received another demonstration. The product was SketchUp (at the time, it belonged to @Last Software). In its digital environment, 3D objects function (more or less) like their physical counterparts do in the real world. If you pull on them, they stretch—it’s as simple as that. When the briefing was over, I was forced to rethink what I considered intuitive. I had come to accept the established modeling protocols as intuition. The two are not the same.

In the last several years, direct modeling—an approach that relies less on our familiarity with protocols—has been enjoying a Renaissance. It’s been around for some time, but SpaceClaim, which publicly launched in 2007, put the spotlight back on it. PTC, a parametric pioneer, is now adding direct-editing tools to Pro/ENGINEER. Siemens PLM Software’s NX with Synchronous Technology and Solid Edge with Synchronous Technology are now in their second incarnations. Not to be outdone, Autodesk has just developed and released Autodesk Inventor Fusion, a technology preview that promises bidirectional exchange between parametric and direct models. SolidWorks and Alibre, two other household names in CAD, have so far resisted this trend, citing a lack of customer demand.

In my view, dismissing direct editing is rather like ignoring the elephant in the room—you can do it for only so long.

In the blogsphere, discussions about direct editing sometimes devolve into semantics—for example, the distinction between direct editing and direct modeling. To add to the confusion, vendors use different terms to describe the same approach. PTC would rather call it explicit modeling.

I propose we move beyond the direct vs. history debate. The shift to direct modeling—or whatever you want to call it—is but a move toward a more natural way of creating 3D geometry. It’s an attempt to shed some of the complexity accumulated over time, a return to a simpler way of working with primitive shapes. So let’s drop the name-calling. Instead, let’s concentrate on developing a modeler that’s immediately understandable to an 11-year-old kid or a tech-phobic octogenarian who’s never used CAD. Then, and only then, should we claim that our technology is intuitive.

Kenneth Wong writes about technology, its innovative use, and its implications. He has written for Computer Graphics World, Cadalyst, Game Developer, and Manufacturing Business Technology. You can follow him on Twitter at KennethWongCAD, or send e-mail about this DOF to [email protected].