RobotX 2026

Weekly Sprint 7 Meeting

April 27, 2026

We are so back (in Notion)

Meeting Overview

Vehicle readiness video prior to finals

</aside>

USV Readiness “Almost” worked (Hopium!!!)
- Gate 1 → Short Buoys
- Gate 2 → Tall Buoys
- OAK-D model works for short buoys but provides incorrect pose values for tall buoys causing us to miss the second gate.
  - Jeffery tried to debug it but we had to pack up cause of rain ;(
STBD oak-d dies very quickly for some unknown reason
- Works just fine again if you restart hardware
- Likely from starting and restarting hardware
Semantic Map (gtbbx_mapping)
- Had to make a number of changes in terms of gate widths and tolerances to allow gate detections
- Modified some gate duplication logic to prevent gates from spamming in the MetaGate list
Behavior Tree (gtbbx_autonomy)
- Regression(Gate shift) detection was too sensitive and caused the gate targets to shift around too much, causing a stale behavior
  - Fixed by changing threshold parameters
- Search behavior works incorrectly for some reason
  - Possibly caused by small turn angles
  - Needs to be improved to rotate “and go closer to gates”
MAVROS Distance values are significantly blown up
- Wouldn’t be surprised if they were in feet instead of meters
- Probably be the reason behind some of the software issues above

Rewrite “detailed” dry test plan and lake test plan for 4/25 Lake Test
- Step-by-step such that it is reproducible by anyone
Blueboat at Sweetwater
- Shakedown BB05 if it works, do gates
- Bring BB04 for software backup
Drone at Waterworks or bust
- Bring BB01 for GPS
- Bring Anker