Documentation

ORCA's second in-water test: leak checks and movement scripts go smoothly until a misaligned Pixhawk and ZED 2i send the sub into the floor, and an overtightened kill switch lets water into the battery compartment.

Finalizing our behavior tree sequencing against the competition task descriptions, cleaning up the workspace, and testing the camera against last year's model.

Tackling the camera lag that dropped us from a target of 720p @ 60 fps down to 0.5 fps, and why we moved our model runtime from ONNX to TensorRT on the Jetson Orin Nano.

Early brainstorming and sketches for CLAW. Rough ideas for now, to be elaborated later.

The full development log of the dropper, from a 38mm rough prototype through five iterations to a final straight-finned, screw-in design — and its accuracy testing.