Before the Test:
Everything in the lab was good before the test. We attempted to hot glue a potential port which could have caused the leak from the pool test on 4/5. We resecured the 1D sonar with zip ties to orient it in the correct direction. I changed the mode of the IMU from indoor mode to absolute mode in order to potentially combat the yaw rotation issue. This ensures that the IMU relies less on its Kallman filter and more on the earth’s magnetic field. When making sure everything was secure Lilly noticed that the balls in the marker dropper had rusted and needed to be removed. The marker dropper is covered in rust and must be replaced. Mechanical needs to decide on a different choice of marker. We also greased the o-ring in hopes of a better seal.
Additionally, Alex, from Pratt media, planned on meeting us at the pool to get some footage.
Test:
We made sure to keep track of the humidity sensor throughout the test to see if water was getting in the robot. We put the robot in and adjusted the buoyancy to counter the lost weight of the markers. As the robot was in the water, the humidity sensor increased linearly, starting from 50% and increased until it was 80%. This happened over the course of between 35-45 minutes suggesting the leak was small and continuous. Before the humidity sensor reached 80, we moved the robot around with the joystick to see if this increased the rate at which the humidity sensor increased (it didn’t). After reaching 80 we took the robot out of the water and found a medium amount of water on the bottom layer of the robot. Since Alex came to film we dried it off and decided to test for 30 more minutes.
Once back in we did joystick control for 15 minutes to collect footage and then we did an autonomous depth hold at the end. The depth hold still had the same yaw issue so this will require further looking into when we have the robot leaks fixed. After we took it out we headed back to the pool keeping track of the humidity in the robot and it steadily decreased on the way back, suggesting the seal was not air tight.
When moving with the joystick we realized that the controls were inverted. The thrust stick yawed the robot and twisting the motion changed the robot’s z position. We fixed this inversion and also inverted the thrust stick such that pushing forward makes the robot sink as this matches the normal control scheme with a joystick like this.
After the Test:
We gave the robot to Vincent and hoped that he fixed it.
Positives:
- Robot still works
- We are done with Pratt footage and should expect a video in the summer which we can modify for our Robosub submission
Issues:
- Robot still leaky
- Yaw issues persist in state
- Mount 1D sonar
- Wire 1D sonar
- Wire servo for marker dropper