PREQUAL COMPLETE!
Today we completed the prequal. We used the same strategy as outlined in the previous update, but with two important changes:
- The code to compute the yaw angle offset between the lane line and robot had some bugs, which were fixed.
- Sometimes the robot wouldn’t correct its yaw angle completely, which led the robot to go near the wall and lose sight of the lane line. This was fixed by adding a fixed angular correction whenever the robot strayed too far from the lane line.
Thus, the robot followed the lane line to the marker, dead reckoned around the marker, and followed the lane line on the way back. This proved to be a reliable, repeatable way to complete the prequal.
Here’s the link to the prequal videos, which have been uploaded to RoboSub’s Box. https://robonation.box.com/s/xkxvrfygs6tw3q0mf56b80ut12b5bowj
We are the 6th team to complete the prequal. I’ve submitted the prequal form to inform RoboSub of our submission.
A few other things:
- The path markers are now negatively buoyant, so they stay on the bottom of the pool.
- We captured a bag file of path marker footage using the bottom mono camera to test HSV filtering. It’s been uploaded to Google Drive.
- The monitor that fell into the pool does not turn on.
- Humidity stayed at 65% throughout the test and there was no water inside the capsule when we took it off afterwards.