3 ms·
I was part of a FIRST team in high school, I think the second year of the competitions. One other friend and I were big into RC cars at the time, and we went i
by feralchimp 15y ago
I was part of a FIRST team in high school, I think the second year of the competitions. One other friend and I were big into RC cars at the time, and we went in with all kinds of designs for drivetrains, oil-damped suspensions, control linkages, etc.
We had a decent design voice (shock/horror none of our oil-damped suspension designs made it), but essentially zero construction was performed by students in the club. Engineers from our partner org built the chassis from whole cloth using laser-cut polycarbonate and other legit machine tools. We didn't even get to watch that happen...the first rev of the mobility platform just sort of showed up one week.
At the time, my friend and I were really disappointed by that. But we stuck around and I'm glad we did, because it took a TON of teamwork and not-infrequent heroism for all of us to use and improve the robot without destroying it.
Software would have allowed far more room for experimentation and individual tinkering, but the group-social stakes would have always been way lower. Our robot literally had cheerleaders by the time we took it to compete in NH. That mix of fragility, anthropomorphism, weirdness, size/heft...can embody weeks of teenage group effort in a way that even a wicked-sweet app just can't.
So merits on both sides, I guess. And hey, app-controlled robots? Boom.
- ja27 15y agoI have mixed feelings about some aspects of FIRST Robotics, mostly about how much kids really learn when they're so focused on winning. (That's actually a core value of FIRST - "What we learn is more important than what we win", but a lot of teams don't really follow that.) FRC (the biggest class of robots) gets all the press, but FIRST LEGO League is a tremendous challenge for elementary and middle school kids. No remote control, completely autonomous operation of the robot. (Well, the students can pick up the robot and run different programs, but there's no joystick drive.) The real physical robots vs. all software / simulation is a classic debate. I heard MIT's Rodney Brooks (of the "build real robots" camp) used to tell prospective grad students something like "it may take you a year longer to graduate with me but you'll have one hell of a demo tape." This year's FRC teams will be getting Microsoft Kinect sensor bars to use. Should be some interesting stuff.