Team 0x27, winners of last year's AVC posted up a video of preliminary testing of their unicycle robot...
Meanwhile, Savage Solder, a recent Robomagellan champ, is performing brilliantly in this test video:
We've heard from Paul B who will be in the Doping Class this year with a high-zoot GPS using a proprietary DGPS network. Team Viator is back again this year, too. Wayne H is back this time on his own with some sophisticated, elegant hardware powering his ECX Torment. His daughter is running her own entry this year and the video looks promising.
My robot? Still not very far along. I didn't plan to change much. What I've actually been able to work on over the last several weeks is:
- Changing the path following algorithm (time to test)
- Tuning the suspension for the ramp (in progress)
- Trying to data log ramp jumps (failed, still working)
- Evaluate a uBlox GPS working (failed, giving up)
- Migrating to offline compilation (failed, giving up)
In addition, the left encoder has been generating suspicious data that suggest wheel slip, but it could also be a software bug or hardware glitch. Error in the encoders ruins the robot's position estimation. I haven't had a chance to even look at position estimation yet.
Because I've not untethered from mbed.org or Google Earth, I require internet access at Boulder Reservoir and haven't sorted that out yet. I've ruined my main battery so another is on the way. A number of details remain to be worked out.
Unfortunately, I'm pretty well out of time due to some important responsibilities requiring my spare time over the next several days. No matter what happens, I will be at the AVC in a little over a week and looking forward to seeing everyone there!
Good luck, man. Fingers crossed for DataBus keep evolving!ReplyDelete
This is incredibly cool! A unicycle is fun but the idea of a unicycle robot is even better.ReplyDelete