Quote:
Originally Posted by BrendanB
From what I saw this weekend 95 had another solid showing at Hartford. I really liked the 6 tote grab out of the landfill!
|
Thank you very much! I was pleased with our alliance making it to semi-finals at such a competitive event, thank you 3146 and 173!
The 6x tote pull is a subset of what we've been working on. In theory, though we have yet to execute on the field, (read: take this with a
huge grain of salt) we could collect 8-10 totes from the landfill and make two 4-high stacks, a 4-high and a 5-high stack, or (very optimistically) two 5-high stacks. Unfortunately this mechanism is very much secondary to our RC arm and thus is less robust, many things have to work out well to mine this many totes. The best we've done thus far is to pull a 4-stack and a 2-stack and cap both. Still, I consider this is a big step up from our 2x2 mining from GSDE. We're now right at 119.8-120.0lbs, which makes it quite difficult to add any more functionality to the totem! All we can do now is practice and refine code.
Speaking of code and practice, a
big thank you to 1519 for helping us with, and inspiring us to implement, an IR sensor array that has been key in making our robot more effective at RC mining. We spent much of our time at Hartford refining the control code for this and practicing using it. Here is a picture of the array installed on our lower-most totem bar. These four sensors allow the robot to automatically square itself to, and center upon, two totes end-on. It has worked well in tele-op, but not as well in auto, yet.
We moved the air cylinders for our 'antennae' to be base of the arm and connect them via fiberglass push-pull rods to the antennae. This moved enough mass down off of the arm to dramatically improve arm performance with the antennae mounted. We also added a small plastic 'wheelie bar' on the front of the robot that lets us pull wheelies routinely when lifting RCs.
Antennae cylinder mounted on the arm shoulder:
Overview shot of the push-rod actuated antennae. The fiberglass push-pull rod is red, the carbon fiber antennae is black.
The geometry for these work, and we've tested it in teleop. We just need a slightly more robust auto code to align properly to grab two RCs in auto. Currently we're brainstorming ways to make this happen.