I decided to take 3 robots and head up to Las Vegas for the LVBots 2007 Spring Event. These included FlatSpat, Borg, and my first linemazer "Black Pearl".
I arrived at Rancho School for the event and started setting up. My problems began with FlatSpat. Since the November PAReX event, I had added a 3rd eye in the form of an LV-Max sonar, to detect free loaders sitting on FlatSpat's blade. I had to relocate the Sharp range sensors to do so, and didn't realize they now fell outside the 10cm size limit. LVBot officials were kind enough to let it slide by. Then I placed FlatSpat on the ring for a test drive and watched it turn in circles like a bot possessed! Now in full panic mode I fired up the laptop and dug out the debug tools. My debug panel showed me the left floor sensor was failing and always saw white, thus the robot was continuously turning away from what it thought was the edge of the ring. Since there are two edge sensors under the blade, I modified the firmware to ignore the left sensor and went for another test. Ahh!, normal search loops! But wait, it's not turning correctly to follow an opponent! Flashbacks of the original turning problem from a few years ago quickly return. I must have started the LV-Max update with an old version of code! A few more quick edits and FlatSpat is ready to go.
All this time they were running the Dead Reckoning event and the Line Follower, so I didn't miss my events. Next was Line Mazer, so I brought out Black Pearl and started. My plan was to use it's default learn mode (speed 2) for the first run, then the default solve mode (speed 2) for the 2nd run. Then with my remaining 2 runs I would dial it up to speed 6 and 7 for my best time. Things were looking good and even my low speed was faster then anything that had run so far, then disaster struck as BP totally missed a branch and headed back to the start area. I aborted the run and tried again with the same results. Even dialing it down to speed 1 didn't work and I was out of runs! Eventually after I returned home, I duplicated the error and determined it to be a combination of logic and over-running the sensors. The fix is quite simple and will be in place before I use Black Pearl next time.
Time for some Mini-Sumo!! LVBots uses a round robin pool and takes the top 4 robots into a single elimination tournament. Using 3 rings we paired up and quickly went through pool play. FlatSpat didn't do as well I would have hoped, only winning 3 of the 8 rounds. Borg on the other hand, assimilated 6 other bots to give it 2nd seeding in the tournament. Borg went up against Happy Jr. and won the first round, but failed to take the match. Invid took down what was left of Borg, giving Borg a 4th place. Not bad for it's 2nd appearance, and I have a few ideas to improve things for the Fall events!
Well it was a long drive up, and an even longer one back with not much to show for it. It always a blast to hang out with fellow bot builders and see their creations and ideas. I learned a few things I might use in my events, and maybe a few thing to do with the robots. Thanks to LVBots and Pololu for hosting a great event!
|
FlatSpat sported a new eye for this event, but limped along for 3 of 8 pool victories |
|
Borg did well in pool play, assimilating 6 other bots, but failed to finish them off, getting a 4th. |
|
Here's my new line maze robot "Black Pearl". BP was the fastest bot on the lines, but that doesn't matter if you can't navigate the maze! |