Here we go again…
So, when we left our intrepid travellers, they had just qualified, and were raring to go for their semi-final run at the early-morning hour of shudder 7am.
After having tested for several hours the previous night, they were confident they could make a run at the gate, with little to no difficulty. Sadly, this was not to be the case. Will they ever make it through the gate? Will they ever be able to make their robot work? Will Dave ever get a haircut? Stay tuned…
But seriously folks, we arrived at the TRANSDEC for our 7am run. This was a result of the order in which we qualified, and since we were the 21st team out of 21…we got to choose our timeslot last. Obviously, we got the only one no one wanted.
But, troopers we are, and get up on time we did. The nice fringe benefit to arriving so early was that unloading and parking was an easy affair, as many of the other competitors were likely still tucked into their nice, warm, soft beds.
As soon as SubmURSA took off for the first run at the gate, things began to go awry. I engaged the switch, it immediately descended and began to rotate to starboard towards the wall of the pool. I killed the run, we reset the program to compensate, but yet again, same thing. This made no sense, considering our testing runs the night before were straight and true, for the most part. So, we decided to try pointing SubmURSA away from the intended heading, hoping it would end up in the right direction. This didn’t work either, as it simply began to rotate in a circle. Puzzled and frustrated, we reset for another run, but I was a bit forceful as I pulled the communications tether off, and the pressure hull came loose from it’s Velcro mount. I spent our precious last few moments frantically trying to reattach the hull, but it was too late. Dave reminded us there was still tomorrow, but still feeling a bit dejected we headed back to the tent.
Somewhat puzzled at this unusual problem that made little sense (the kind we are used to having), particularly in light of our hard work the night before, the only thing we could think to pin it on was a compass issue. This was to be expected, given this was the first year for these devices. Veselin began running some tests on the compass in various configurations to see if he could get consistent readings. Scott went to work physically mounting the pressure hull to the frame with bolts, since the temporary velcro solution was obviously proving to be unreliable when it was needed most. Meanwhile, I booked a practice run for later in the practice side of the pool.
Some preliminary tests seemed to indicate the compass had something of a ‘dead spot’ where the heading seemed to go off. This could be compensated for by some more calibration, although neither Veselin nor I was convinced this was definitely the problem. There were several parameters in the heading controller that could be tweaked to account for strange readings, and we did have the option of using the gyro to keep the heading in check, so Veselin set to work on implementing these changes.
After several tests on the bench and in the dolphin pool, our practice run came up, and we were able to do some more testing under competition conditions. While doing gate runs with the tether attached, we discovered that a very large correction was being applied to the right, while the robot continued veering left. Once we got SubmURSA out of the pool, we attempted to check the systems, while I connected the current sense lines to the motor controllers so we could see if they were turning as directed. Once we had those connected, the values seemed to correspond to the loads on the thrusters, however one thruster that was pulling current was not actually spinning. Scott stuck his finger in the shroud to give the propeller a kick, and it spun to life with barely enough time for him to pull his hand clear. Interestingly enough, our problems that morning could have been explained by a simple stuck thruster. This made sense, as I’ve seen the same behavior when a motor controller failed in the past.
Knowing that, we re-positioned the thrusters, placing one of the less-used thrusters in place of the one that was sticking. I had tried to book us another practice run, but the list was full. I got our team a spot on the waiting list, knowing it was unlikely we’d get in the pool again today, but it was worth a shot.
So we tooled around for the rest of the day – Veselin still working on understanding the compass, myself making tweaks to the motor controllers and main board, Lee figured he could solve the problems with the sonar board, and began working on it again in the off chance we could use it this week, and Scott worked at securing the pressure hull to the frame.
We didn’t really expect to get in the practice pool again, as it was booked solid and we were 4th in line if a spot opened up. But, only a half-hour before the last slot, one of the judges came around and said the last slot had opened up if we wanted it. Briefly surveying the team, I announced that yes, we wanted it.
So, we quickly reassembled SubmURSA and began some preliminary testing in the dolphin pool. Our practice spot was only a few minutes away, and I still had to attend the leader’s meeting to choose our semi-finals spot for tomorrow. Giving the order to get SubmURSA to the practice pool, I went to the meeting and based on our static judging score, was able to secure a 9:40am semi-final time, much better than our previous slot.
At this point it was time to take to the practice pool, and we did a couple more runs with the new configuration. Things seemed to be working reasonably well, and we were able to get through the gate more than once. Too bad we couldn’t do that in the morning.
We packed up and headed back to the hotel. After some dinner with our advisor, Dr. Moussa, and a long walk back to the hotel (since our shuttle bus left without us), we were all too tired to do anything else, so we turned in for the night, ready to give it another go tomorrow. To the gate!!