Droid building

My brother and I decided to build a droid in late 2006. I had doodled some robots during the past two years after I re-discovered David L. Heiserman’s 1970s book “How to Build Your Own Self-Programming Robot.” I say Heiserman’s book, but it also was very much a Tab book, one I had come across in a library in the 1980s. Tab books in both decades were filled with nourishment for hobbyists of all stripes.

I showed the book to my brother, an engineer, and asked whether he thought it was realistic to build Heiserman’s creation in the 21^st century. He quickly pointed out that not only were many of the parts specified no longer manufactured and thus hard to find, but also that today’s microcontrollers could do more in much less space. Microcontrollers seemed to take away from the build-it-yourself aspect, but then Heiserman called for an 8085 chip (price from Jameco was about $5 in 2006) to be used as the core of his robot; manufactured parts of some kind would be required. The only question was what degree of manufactured parts would be acceptable.

I never did like, as a child, attending “make it yourself” projects such as “make paper” with instructions such as this, “First, take some paper. Then tear it up.” To a child interested in how things work, that sort of move is contrary to making paper. I wanted to know about pulp. But Heiserman’s use of an 8085 for processing power doesn’t really fall into the same category. And how far should one drill down to the essentials to be satisfied that one is truly creating something? If an 8085 is too much of a manufactured black box, imagine the cost and space of the equivalent discrete components. And why stop there? Should the purist assemble his own transistors? Buy raw silicon, dope it and cook it? Mine silicon and boron? Using an 8085 in a homebuilt robot is something better than tearing up manufactured paper to “make paper.” Perhaps microcontrollers should be treated the same way.

At any rate, I was doodling robot ideas through 2004, 2005 and 2006. Yet my sketches kept resembling R2-D2. At some point, I decided not to fight the urge.

But which R2 to build? There are two kinds: Real and ... something other than real. There are many real R2s, and not one of them is a robot. There are radio-controlled models, computer-generated models and the Kenny Baker model. The radio-controlled models aren’t real robots, either. A real robot is autonomous, independent of action from its human masters. Heiserman gave his creation what he called “machine intelligence,” a term he suggested had more legitimacy than “artificial intelligence,” which after all implies “fake” intelligence.

Yet the other R2 does a better job of capturing the imagination. It is the imagined version, the idea of R2-D2. The fake R2.

In the Jan. 8, 2009, posting to his blog The Technium, Kevin Kelly mentioned the subculture of movie prop copiers as part of a larger discussion of modern “as if” desires. Kelly, editor emeritus of Wired magazine, talked in the entry about the creation of things as imitations of other things but which gain so many layers of depth they become outright “things” in themselves:

I have a friend, Adam Savage, co-host of the Mythbusters, whose hobby is creating models of famous Hollywood props. (He is not alone in this obsession. There is an entire subculture of prop copiers.) Adam spent several years working on his off-hours to make an exact copy of the Maltese Falcon which starred in the movie of the same name. The prop in the movie is a contrived imaginary sculpture that looks nothing like the original gem, but Adam wanted to re-create the prop and not the original. So he spent an insane number of hours tracking down photographs of the prop, scanning them, sculpting, and eventually casting a duplicate of the "original" prop. He was obsessively making an original copy of a fake, because the fake (the movie prop) was itself hyper-real; it was no longer "as if" but something in itself.

In late 2004, ahead of the Celebration III convention planned for the following spring in Indianapolis, I found the R2-D2 Builders Group. It is a Yahoo group dedicated to building replicas of R2-D2 and similar astromechs. (There are several group related to droid-building, one of which is more general than R2-D2 and called, appropriately enough, Astromechs. Astromechs has an alternate Web site, astromech.net.) The R2 group includes complete dimensions for all of R2’s visible parts, photos and hundreds of messages related to replica droid-building.

A replica R2 begins as an imitation of something that itself was never real but only a representation of reality. Yet a visit to a “Star Wars” convention’s droid room will show an array of real-fake droids that have become more than movie prop imitations. Hobbyist droid builders tend to add features to their replicas – radio control, moving periscopes, dome lights, robotic arms, moving periscopes and even projectors that stand in for holoprojectors. The features make each droid something more than a mere replica, often a robot (parabot actually) in its own right, with its resemblance to a movie prop ending past skin depth.

The add-ons, I think, are part of what give droid-builders so much enjoyment. Problem-solving is another. And seeing a piece of a much-loved movie in one's living room must be part of the equation. The builders have been active for a decade, and the Yahoo group has progressed from very educated, and sometimes laser-measurement-device-assisted, guesses about proper dimensions to actual dimensions taken from an Industrial Light & Magic droid, which some club members were allowed to measure after a sympathetic droid expert learned of the group. Club members have produced some remarkable units. Follow the links above to see.

As for my own real-fake R2, it now has a real beginning:

It is an A&A Frame (named for its designers) made of PVC. The PVC is lightweight but sturdy. It was designed to hold shoulder motors, a dome motor and more. It was designed to be a less expensive alternative to aluminum. The frame you see in the photograph was purchased a couple of years ago as part of a “parts run.” From time to time, club members will pool their resources and place a single order with a machine shop or other vendor to produce droid parts. A&A – Alex Kung and Andy Schwartz – designed this frame, took out bids and let members get in on the order. After a two-year delay, I took the frame pieces out of the basement and began fitting, gluing and screwing the pieces together.

Spaces for the horizontal utility arms are clearly visible in the photograph. Boxes for the vertical arms also are part of the design. The center vent ports are unmistakable, as is the power coupling space.

Next steps include sanding down the glued portions and attaching skins.

It’s a start. And whether I’m building a real-fake or something greater – I hope it is the latter – it is sure to be enjoyable.