Design: invented or discovered?

“Dave, it’s like your taking all the mechanisms from the last four years worth of robots and sticking them all on to one robot.”

“Max, this is our ‘bag of tricks’. I’m taking everything we’ve learned from the last fifteen years and sticking it on one robot.”

I’m not supposed to say anything about the design of the robot we’re building, even on a blog no one from FIRST reads. But suffice it to say that my team seems to be slipping into old design habits, building a robot the way my family has leftovers for dinner.

The idea is to follow a decision tree of all the possible strategies, then designs, then specific fabrication details. You go from vague ideas (“awesome game piece manipulator”) to concrete designs (“elbow and wrist arm that grabs the inside of the tube”) on down to mundane details (mounting motors in just the right place in the slot to make the chain taut). The theory is that choosing the best option at each branch will produce the optimal robot. It’s a theory that I have to go with, because if I reject it, there is no method to locate the optimal robot in the decision tree that is both systematic and logical. (Make all the worst choices and get the best bot! …..not.)

Leaving that aside, your mileage varies based on your ability to see as many possible paths as you can, and then to decide which one is optimal. This decision rests on your understanding of the task (what you think you are required to do), and your fabrication abilities and limitations, and your understanding of your own fabrication abilities and limitations. Which means the decision tree is different for everyone because that last bit – knowing you can build a precise part on a tight schedule with high schoolers working whatever level of machinery you have – is different for everyone. Some teams have hacksaws and rivets; some teams have bandsaws and a student welder; some teams have CNC (computerized) milling machines and professional welders. My hometown team falls in the middle category; my Boston team in the first. There are plenty of teams in the third, however, with professional engineers from big name companies. And each has a completely different bag of tricks.

Todd summarizes: “There is no such thing as a perfect design. Every design is a bucketload of compromises.” It’s about finding the compromises you can live with.

This line of thought awoke the Apple fanboy in me, which remembered back to an interview with their chief designer, Jonathan Ive. He said that an Apple product is “undesigned. You don’t ask ‘why does it look that way’. Why would it look like anything else?”

This is the antithesis of the decision tree model. The design did not emerge fully formed. You can always say why it would look like something else by taking a different decision branch. It’s not obvious that you should have this function key be here; in fact I think the one for the Dashboard is a waste considering I never use it. (But repurposing it would require installing a third party modification, so I’m not going to muck with it.)

That said, some of the design ideas on my laptop did seem to arise from epiphanies, outside a decision tree. Like using magnets to connect the power cable and keep the screen closed, even when techies are afraid to get magnets anywhere near electronic storage. But the result is a power cord that won’t pull my laptop off the table when someone trips over the cord, and the removal of a mechanical latch to get broken because I’d fidget with it. These sorts of design ideas are obvious only in hindsight, because they remove complexity you thought was necessary. Trackpads have to have a tracking surface and a button, right? Wrong: make the whole surface a button and click anywhere.

On the software side, I remember trying to make a numbered list in Pages, Apple’s word processing software. I kept looking for a button to activate a numbered list, which I expected to be next to a bulleted list button, just like the ubiquitous MS Word. No dice. I had to go to the help system – oddly good at addressing specific technical queries –  to find there was no such setting. Just start typing a numbered list, and the program will make the next number for you. I would have never thought of that because it’s simpler. But now, I ask Word, Why would it be any other way?

FIRST, it seems, is quite intent on providing a narrowly defined challenge devoid of loopholes or creative, off the wall strategies – the sort that arise from spontaneous epiphanies. Last year, one of our team members had a crazy idea that essentially broke the game, and even we shied away from it. Despite a revised, more stringent rule seemingly forbidding it, another team used the same idea and won six regional competitions and second place at the championship. I celebrated that the game breaker lost to a team that did the job as intended and did it well, but it was really because the breaker’s ally screwed up.

This year, the same team member had another idea regarding the design of the minibot, a completely separate robot built from a new (to us) construction system, that is intended to climb up a pole as fast as possible. The epiphany was to not give the minibot motors, battery, controller, and so on at all, but rather propel a small piece of metal upwards with force provided by the main bot. You know that carnival game with the big mallet and you send the piece of metal as high up as you can? Same idea.

This time, FIRST put a nix on the game breaker fast. Three days in, the idea is dead:

<G19> MINIBOTS must … move up the POST solely through electric energy provided after DEPLOYMENT by the permitted, unaltered battery and converted to mechanical energy by the permitted unaltered motors (and associated, appropriate circuitry).

In it’s place, we have a game that relies of programming, precision, and an inflexible strategy. A game that is sterile with no room for creativity. The worst of both worlds: no epiphanies, and no interesting places to fork in the decision tree. The following is not an enlightened insight but a defeated statement of boring fact: Of course our robot does that. What else would it do?


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out / Change )

Twitter picture

You are commenting using your Twitter account. Log Out / Change )

Facebook photo

You are commenting using your Facebook account. Log Out / Change )

Google+ photo

You are commenting using your Google+ account. Log Out / Change )

Connecting to %s

%d bloggers like this: