Dude, Your CAD's Showing

AutoCAD turned 25 last week. That's an awful lot of legacy to shoulder, especially for a product so bound to change as a software package. The fact that it's still around (how many other 25-year-old programs still exist with their original name?) is a testament to a few things: being first to the party, being very good at doing something a lot of people needed, and being really hard to learn. Architecture and engineering firms are full of experienced AutoCAD drivers who once suffered through an interface that's excruciatingly pedantic by current standards, and built a career on having done so. The only alternative was the even more laborious and mistake-intolerant science of hand drafting.

They worked on through the 80s and 90s, confident that they were using the state-of-the-art, and glad to have been freed from the unforgiving world of Maylines and powered erasers that their seniors knew. And then 3D CAD showed up, expanding software's role from a drafting-board replacement to a studio necessity. From the mid-90s to present day, we were presented a bewildering assortment of options, similar enough to make choosing the right one a chore, but different enough to inspire deep-seated assumptions in their users.

Although there's a common saying in the Industrial Design community that "CAD is just another tool," and a truly skilled designer uses whatever tool is necessary and appropriate, it's quietly understood by many that your CAD package speaks forcefully about what kind of designer you are. Anyone familiar with the field will already be aware of a sort of crude continuum of "real ID": parametric solid modelers (Pro/E, SolidWorks, CATIA, Unigraphics) are for engineers, and by extension, designers who are really just engineers that don't do math; surface modelers (Alias and Rhino) are for "creatives" who can't be bothered with manufacturability; and if you're a Real Designer, you can do everything that needs doing with a bin of pens and markers. It's an invented hierarchy that leads to endless pissing contests about sketching skills, but considerably fewer about modeling skills: the stereotypical hot Alias jockey is still more self-effacing technician than rock star, and Pro/E and SolidWorks don't even really have "jockeys," just users.

Certain industries have long associations that lead to these perceptions, like Alias for the competitive, art-driven auto and cellphone industries, and CATIA for the wonkier world of aerospace. The visceral nature of hand sketching probably deserves some of the blame too. Watching someone sketch, after all, is fascinating. Pulling CVs in a surface model is somewhat less so, but still has a kind of sci-fi demigod quality to it. Parametric solid modeling, by contrast, with its pop-up dimension boxes and methodical ordering of features has little appeal to the non-user. It's a bit like watching a flower grow: an amazing process that will probably produce something beautiful...but couldn't I just come back and see it when it's done?

The CAD-enabled designer, then, is defined by his or her package of expertise to a greater degree than we'd like to admit. I'm sure I'm not the only one who's scanned down through a job posting, looking for a familiar software title before letting myself explore the other requirements and duties more carefully. Searchable job databases make this even simpler: just filter out everything with "Rhino" in the body, and go from there. There are tacit assumptions among many employers that a) the mindset associated with the desired CAD package will naturally bring the right applicants in, b) it'd be too much time and money to train an otherwise ideal candidate in a new program anyway, and c) there are enough hungry designers out there that such specificity is permissible.

Unfortunately this makes it easy for designers to get pigeon-holed, to the detriment of all involved. The old trope is true: CAD, like sketching and Sculpey and foam carving, is just a tool, and like all tools, it influences how you design. One client during my freelance days--an architect since the 1960s--talked about the times when just by looking at a product he could identify what software had been used to create it. Boxy, blobby, sharp, linear...they all had a culprit. That's going away with better interfaces and newer releases; some say it's already gone. The belief lingers though, at this point more a function of the assumptions of designers than the limitations of the media. Sinuous, edgy sporting goods don't get designed in CATIA, but this has less to do with the limitations of CATIA than the fact that it's expensive, complicated, and attracts those who need its analysis tools and high-precision geometry control. There's little incentive to learn something so daunting when your industry of choice is set up to take 2D sketches and renderings and throw them at engineers...and with all that effort going into improving sketch ability, who's got the time?

Designers tend to pledge allegiance to the tools they put the most effort into mastering, and often extend that allegiance to their hiring practices, seeking those juniors already possessing similar proficiencies to their own. In some cases this is finally curbed pragmatically: hardly any ID programs teach AutoCAD anymore, because it's really quite ill-suited. In most cases, though, a studio sticks to one or two packages, often defined by what the seniors learned, and new hires are filtered by their knowledge of it. Change comes slowly, in the form of a major client demanding a particular format, or a shift in management. Very occasionally, a studio that's truly committed to self-assessment will sit down and examine the alternatives, and decide it's worth learning a new program because it simply suits their needs better. In reality, the time it takes to get a studio up and running on a new piece of software can be measured in weeks, especially if professional training is used, making it a no-brainer as a business investment if the package is truly more appropriate.

In a perfect world, every designer would have every tool at their disposal, and be able to select the most applicable; or else only one tool would be necessary for the majority of design problems. But there's no Category Killer in 3D CAD, like Adobe Creative Suite for 2D or MS Word for word processing. The balance of rapid, intuitive interface versus industry-quality precision means no one package can cover even most of the bases, and a fear of steep learning curves discourages switching packages. SolidWorks--probably the CAD software trying hardest to be all things to all product developers--makes much marketing hay of its file conversion and importation abilities for this reason. For the designer looking to get hired though, the options are scarce. Either learn all the tools you can, or submit to the stereotypes. Perhaps in another 25 years we'll no longer have to make that bargain.