Despite all the innovations in design software since its inception (2D to 3D, parametric solid modeling, Windows-based UI), CAD remains difficult to learn and use.
Computer-aided design (CAD) has come a long way in the many years we have been involved with it. In its ancient history, drawing lines on a screen kept a mainframe busy. Since then, CAD has freed us from the drafting board. We can rejoice at not having to practice lettering, use 30/60 triangles, lead holders and electric erasers. CAD has given us precision and perfection and has rocketed us into space (3D), color (renderings) and movement (animation). Is CAD a done deal, though? Is it full of every possible feature, granted every capability, as is the consensus of CAD vendors? Indeed, we have every conceivable way to make a perfect circle (point and radius, 3 points, center and point, center and tangent, tangent to 3 entities …), including computers capacious enough to house entire aircraft and the ability to render them better than a camera could. Still, the promise of computers aiding design is still largely unfulfilled. CAD still has a hard time with creating initial concepts (sketching), imagining the perfect shape (optimization), managing organic and natural shapes (for which Euclidean geometry is inadequate), and dealing with discontinuous data (from scanners and sensors).
Why Is This So Hard?
I had to build a little free library. I had too many books. I feared my neighbors had too few. The pandemic has kept the public libraries closed for over a year. It was to be no more than a wood box with two shelves, a sloped, shingled roof, and a see-through door. I had a several CAD programs I wanted to try for the purpose of review, so I spun one up (one that was, by all standards, dead easy to use) and started sketching a part. The initial profile shape was super easy, but when I had to give it an exact dimension, I froze. There was no hint of how to enter the dimension in the program. I looked it up. Next, the part had to be rotated to be vertical. How to do so was not apparent. Documentation was checked. Google was questioned. A rabbit hole opened, and half an hour later, I found myself watching a YouTube video on something far more interesting than “how to start modeling in [CAD product that shall not be named].” A precious Saturday was ticking away, and I was no closer to sharing my books. I closed the lid of my laptop, put on my shop apron and started cutting wood—with nothing to guide me but the crudest of pencil sketches.
Given a multiple-decade background that includes much preaching of the wisdom of designing before building, going into the shop without a fully detailed CAD model was a somewhat shameful act. Other projects were undertaken with detailed parts that had dimensions, making assembly a breeze, with all the confidence that every piece was the right size and everything would snap together. That was true for a little library, for an automobile, or for the Mount Everest in our world—the Boeing 777—the first major aircraft to be completely and totally modeled in 3D (CATIA). The 777’s first flight was the high point for all CAD geeks.
Was I being lazy? Impatient? Shortsighted? Indeed, some of the shapes I cut did not fit together and had to be improvised. The very mistakes that CAD would have avoided.
But in retrospect, the time and material saved from having built with electrons rather than wood would not have made up for the two to three days spent learning how to use new CAD software and another day or two spent doing the modeling and detailing.
All told, the time learning the CAD software would have negated any value gained from using it. Two to three days learning is two to three days I would never get back. Two or three days is only a low estimate. When I taught AutoCAD, it took week-long beginning, intermediate and advanced classes to learn all the commands. In industry, I knew it took six months to a year to achieve true proficiency.
Take any CAD program, no matter how easy to use it may claim to be, and within minutes, you are bogged down, unable to add a feature or make a modification … and there you are, looking through the documentation, wishing you had taken a class or at least sat through a tutorial. You go down one road after another until, finally, you pull over and ask for directions.
Though progress has been made with ease of use through the ages, CAD is still not an application you can just pick up and use productively. Starting out on any CAD program is a struggle. The menus, mouse and screen don’t cooperate. The commands you need are called something different or can’t be found.
By comparison, almost every other software we use is just pick-up-and-use easy. No training is needed. No documentation is necessary. We all learned word processing without taking classes. The same is true with email. We can all use spreadsheets, though some of the more esoteric functions may have us looking for help. Even other professional software, like Photoshop, can be bumbled through for most of the basic functions.
But CAD stops you dead in your tracks. You cannot create the most rudimentary geometry—much less a shape—without some sort of education. Having mastered one CAD program is little help in mastering the next. Sure, you can eke out a circle for an extrusion, but you will get snagged trying to join, subtract or intersect. The controls are not simple—the methods and terminology not common.
There have been strides made in ease of use, to be sure. Ashlar-Vellum was able to intuit where lines and arcs were to end because of an awareness of previously constructed geometry. As you were sketching, a line segment would automatically become parallel to another line, or find a midpoint, or be perpendicular … like the program was reading your mind. The first time I saw Solid Edge, I was overjoyed to see commands and icons adapt to offer what I would need next. SketchUp was to become the most popular CAD program in the world not just because of its price ($0), but also its friendly interface. Recently, I leapt at the chance to try Shapr3D, the most modern of all solid modeling programs. Shapr3D’s interface is like a good waiter—out of sight except when needed.
Still, with each CAD program I have been able to get only so far—and never far enough.
The Challenge
Shapr3D was the last CAD program to sail into view under the ease-of-use flag. It’s been a while. Ashlar Vellum’s ease-of-use innovation, the Drafting Assistant, was introduced in the previous century. Other CAD software vendors seem to have abandoned ease of use as a goal. And why shouldn’t they? Once a CAD program is established, ease of use is no longer a consideration. Users know how to use their programs. Millions of AutoCAD and SOLIDWORKS customers may have forgotten the pain of learning a complex system because it was a shared and common pain. It was to be expected—a rite of passage, so why dwell on it? Ease of use could be considered when users must learn a new CAD program, but how often does that happen?
But we forget that ease of use comes into play not just when a new program is to be learned. Ease of use lets all users—new or veteran—design faster and more efficiently. Because using CAD is the expectation—almost a law in our industry—not using CAD, as slow and inefficient as it is, is not an option. Only when we are free to choose whether or not to use CAD, as I was with my little free library, is the the faulty premise of computer-aided design laid bare.
The future of CAD belongs to an innovator who can complete the picture that innovator companies like Ashlar-Vellum, Solid Edge, SketchUp and Shapr3D started to paint. For the computer to anticipate where we are going, be one move ahead of us, offer the right tool at the right time (like an operating room nurse to the surgeon), suggest the fasteners needed—and place them, follow house design rules, know degrees of freedom from recognizing joints you have picked or created, jump ahead to suggest using a similar part that already exists (background shape search, like Google) completes our queries … shouldn’t that be what a CAD vendor should be doing? Providing big innovation that matters, rather than insisting that their job is done, that their products are robust and mature, that there is nothing left to add—except for a half dozen items from old wish lists that they are finally getting around to?