A friend and I decided to enter into the robotics competition that's held every year in Boulder, CO during a certain robotics exposition. The competition is fairly straightforward. Design a robot that will navigate and get as far as it can down a long stretch of obstacle course. The robot that can get the farthest out in the least amount of time wins. The first prize last year was a mere $300, so obviously money is not the incentive for more and more people to come from near and far to attend this competion. No, no no... 'tis the sheer adrenaline rush of watching numorous robots struggle in vain to reach the unattainable end. (nobody's ever made it all the way through the obstacle course - but we're still trying). So, 2002 was our first year to attend this fun event and we got 4th place out of 10. Not bad, I'd say. We may have walked away empty-handed, but our spirits were fueled with a greater aspiration to win that first place for next year. We'll get'em. We'll make 'em pay...
So, what was AutoREKK like? What made it tick and grind? Many of the parts that were used for AutoREKK were later cannibalized to be used in Jander. It used a pair of Six speed motors with a pair of tires from a remote control airplane. Its chassis was all-alloy. The motors were powered by a pulse-width-modulated line driver which was told what to do by a logic circuit, which obeyed a programmable logic controller (Basic Stamp 2S). There were 'antennae-like' wall detectors made from coat hangers, LED line followers and a pair of infra red proximity detectors.
The approximate cost was more or less as follows:
Don't forget to factor in all the hours we had to rebuild and redesign this thing through more than six completely different design overhauls. What we had in the end was very different from what we had in mind at the beginning.
Through this experience, we learned many valuable lessons. For example, we had no idea that the obstacle course was going to be so challenging. There were holes, carpet, inclines, boltheads and even a pit filled with ping pong balls that our robot was expected to navigate through. I think we did pretty well though, considering our lack of experience up to that point.
So, this year, my friend and I are developing our own independent projects. I haven't given my robot a name yet, though its probably going to be something like "darth taxi" or something along those lines since I'm going to stick to the Taxi motif each year. For now, it's just called "Jander"
Valuable lessons learned:
Treads
We both agreed that treads were definitely the way to go in this competition. Wheels, unless they're really big, are not going to give the robot the stability it needs to traverse holes and bolts. They also don't allow the robot to get enough surface contact to scale inclines
Clearance
High clearance is another issue that needs to be addressed. If you're trying to go across the field of bolts and have a very low and wide chassis, odds are you're going to get stuck.
Appendages
You know that feeling when unwaxed dental floss gets stuck between your teeth? That's how the robot feels when he goes over rug and has annoying line-following diodes get caught in the carpet fibers. Ugh! So, a good idea to avoid this is just simply to pull out all those lousy teeth! That's right, so what if I can't follow lines... big deal! I'll just detect walls as well as pitfalls, and I think I'll be ok. Nah really... I'm serious here!
Flare
At the end of the competition, there's a 'best in show' ritual which is free to attend. Basically, your robot just needs to be a crowd pleaser and you win. I think the bright Taxi motif helped this, but the robot could have used a few more flashy lights. Something to keep in mind.
So, wish us luck for next year!