Sunday, May 5, 2013

What I Do

(Other Than This and The Nerdy Stuff I Write About)

Dumbasscience: Science is not always smart. Oh no sir-ree-Bob. Sometimes it is just plain stupid. That's okay, I won't judge. I'll just point it out and mock it mercilessly. These articles will contain rants from history and breaking news where scientists are pursuing through-and-through bad ideas. From the 'flying your car to work' pipe-dream, to various plans of how to forcibly reverse global warming, if it's technological and perilous to rationalists, it's Dumbasscience.

Over the past year and a half, I've been working at an awesome, steady job in a technical field that is moderately controversial these days. Due to disclosure agreements and the fact that right now I'm a contractor in my position, I can't specify my employer. But I can talk with you a little bit about what I do: 3D printing. Its controversy over the past three years had been almost purely driven by ignorance regarding the technology itself and the products 3D printing can produce. Which makes it a perfect cross between "A Day in the Nerdery" and "Dumbasscience".

It's worth noting that I work specifically as a Rapid Prototype Specialist, making form models using Zcorp printers from 3D Systems. These machines produce unique products that can yield full color parts equal to early inkjet printers across an entire three dimensional model, but they almost certainly aren't the ones you've heard of, if any. No, the popular 3D printers in the headlines these days are usually either the latest from Makerbot or Objet printers by Stratasys. I work with Objet printers in my shop as well, though I'm not the primary operator associated with them, and I've researched and prepared to work with Makerbots for our purposes here at work, so I'm fairly qualified to speak on them. In a word, Makerbots are inferior machines that take longer to make lower fidelity parts than almost any other product in the field, and they're held together by Makerbot's outrageously aggressive PR campaign. Objet machines, on the other hand, are some of the most accurate 3D printers on the market, yielding plastic parts of several different materials in fixed compositions.

What Wired Magazine Thinks I Do

Every time I see a new article come up on or in some other even more mainstream media, I choke back a little bile in my mouth. Almost every article written for mainstream audiences takes a stance that it has to explain what 3D printing is by comparing it to Star Trek: Next Generation's replicator technology. Really? First of all, I must've missed that moment we slipped into an alternate dimension when plot-expedient technology from ST:NG became common parlance? You couldn't think of anything more relevant to everyday readers? More importantly, however is that if a person is ignorant enough that the replicator-allusion is at all helpful to them, they'll get the exact wrong impression of what a 3D printer is. It gives them the impression of a physical object duplicator, implicitly promising scanning, processing, functionality, and fidelity. All 3D printers are based on realizing a form of a model based on the loaded materials, but that's as far as the analogy to replicator technology goes. Scanning in the 3D printing field is crude and superficial to the point of being only a novelty for rough approximations of a model. Processing is a very specialist-intensive process and requires a large amount of finesse from the printer's operator. 3D printers don't reproduce mechanical functionality unless that functionality is extremely basic and scrupulously programmed by the operator, anything as simple as a threaded piston requires a lot of careful engineering, which means you can forget about the requisite dumb question "can you 3D print a 3D printer?".

No. Just, no.

So what is a much better shorthand explanation for 3D printers? They work by printing hundreds of thin cross-sections of a model in a stack to approximate the model's overall shape as far as layer thickness and accuracy will allow. It produces things of a pre-determined material that merely mimics the shape and form, but not necessarily the function, of the part. Zcorp machines even use a roller for distributing a fine, flour-like plaster powder across the build layer. Other machines distribute resins and use lasers to 'cook' the shape into place. So here's the skinny: 3D printers are basically advanced cookie-cutters for making physical objects, with limitations comparable to that comparison. There. A simple, clear explanation that doesn't require the audience be Trekkies nor burden their impression of the technology with overladen imagery.

What the President Thinks I Do

In recent months, various government officials have made passing comments at the potential of 3D printed firearms and used that specter as the pretext to opening up dialogue to control these machines. Now, I'm not getting political with this, I promise. I'm just going to point out that such discussions of 3D printed firearms fundamentally misunderstands both the printing technology and the technology of firearms themselves. As far as I've found, the most successful use of a 3D printed part in a firearm is right here. As you can see, the 3D printed part jammed on the first shot and broke altogether after the sixth. And the only part of that gun that was printed was the lower receiver, which is nothing more than the part which connects the magazine to the handle, stock, and firing mechanism. So no actual moving or stress-bearing parts were fabricated in this manner, and yet it still broke before exhausting its first clip. Experiments to 3D print more elaborate parts will inevitably result in results comparable to ammunition spiking--which at best just destroys the gun and at worst maims or kills the shooter.

And that's before you call into question the cost of 3D printing at this fundamentally inadequate level. These weapon tests are being proposed and conducted for Stratasys Objet printers, which run $150,000 on the low end and easily cost tens of thousands of dollars to supply and maintain in a year of moderate to heavy use. The tagline that a kid could print this up in their living room must presume that said kid is Jaden Pinkett Smith and can blow a house-worth of dough to make an unregistered gun that is going to be less reliable than a Wal-Mart rifle he could've bought for a couple hundred bucks and some paperwork. And that price tag I've estimated above is what 3D gun-proponent Cody Wilson identifies as 'getting cheaper' when he stresses how accessible 3D printers are becoming.

Now, 3D printed guns will eventually become a viable production method, but only after many years of people risking their lives testing obviously unsafe weapons, and even then hobby gun makers are going to be at the mercy of how fastidiously they maintain their printers. Because one quality all 3D printers share is that maintenance and repair issues often impair part quality without impairing build ability--meaning that before your machine actually breaks to the point of inoperation, you're very likely to produce a part that is just close enough to spec to get you into trouble. The shrapnel kind of trouble, that is.

What my In-laws Think I Do

Here I'm using in-laws as shorthand for anybody outside the Rapid Prototyping field. I can't speak for all other 3D printing specialist out there, but I find that once I explain the technology and its innovations to someone, I can't really stress that the job is taxing in anyway. It doesn't matter that being a full-time rapid prototyping specialist requires cross-discipline finesse involving simple mechanics, engineering, 3D design, physical modeling skills, and working in conditions that vary from physically irritating to hazardous to one's health. The second I bemoan being exhausted from work I get treated like George Jetson wailing about how sore his button-pushing finger is after a long six-hour day of work.

What Designers Think I Do

In my discipline amongst 3D printers, I have to coordinate with designers as I help realize their computer-aided designs in a physical prototype. Some designers are better at keeping their work printable than others, but almost everyone on the front-end of my workflow seems to think that once they send me their data--no matter how bad and in need of tweaking--I just hit a button, sit back, and have Miller time until my easy-bake prototype machine poops out their part a couple of hours later.

What I Feel Like I Do

As I mentioned above, I work with designers, and they have their own timelines and objectives based on what's being evaluated and when. But once the requests filter down to me my timeline and my objective is always the same: as fast as possible and as good as possible. Though oftentimes it's needed faster or better than possible, so I need to make accommodations for that, too. The quality of the data and parameters for the models I build, however, vary terribly from clean and specific to vague and fragmented. So on any given day, my machine-battling rage undulates between frustration at fixing other people's work to trying to pound a square peg into a digital hole.

What I Actually Do

3D printing is a fantastic and sophisticated technology. It involves layers of technical rigor, procedural endurance working with 3D software, and artistic competence. I can spend one day mostly seated at my desk doing data and standing at my worktable polishing finished models, or on another day I'll trek back and forth across my work's entire campus several times, lift 40-pound kegs of printing material, and push and pull my way behind 200-pound machines to examine them for cleaning and repairs. It's a technical job that requires a good back, delicate hands, and an artist's eyes.

Fortunately, I have an extra of each in my desk at work.

"3D printers are basically advanced cookie-cutters for making physical objects, with limitations comparable to that comparison."

No comments:

Post a Comment