Talk:DEWBOT XI Midknight Mayhem
From DEW Robotics
Post-Mortem
Contents
What went well
- Drive team communications are improving
- General robot reliability very good
- Stacking good when we followed the process
- Scouting insights between matches
- Alliance Selection
- Alliance learning and improvement during eliminations
- Dodge ball (we won this at least)
- Eric & Dana able to intake upright RC
Issues
- RCs lost in early Qualification matches due to RC being pushed back while loading litter
- (2) RCs rolled off stacks after scoring in Finals 1 (cost us 60 points and the match)
- Robot tipped over in Finals 3 (cost us the match and the win)
- The intake wheel drive coupling key drops out
- Massively inexperienced coach
- RC Burgling reliability was lower than expected
- Auto alignment was not working
- We did not use 3-tote autonomous due to know unreliability
Improvements
- More driver practice
- 2 white boards needed - pit & tactics
- Fix 3-tote
Analysis
Video analysis clearly shows that the "Robot don't tip" software did not engage. Software log confirms this.
Based on testing, it appears that occasionally the NAVx does not initialize. Software was modified to check initialization and if this has not been successful, reinitialize the NAVx.
This fault potentially explains some other failures, particularly a 3-tote autonomous failure during an Archimedes qualifier.
Loss of RCs (video provided): path is less clear here, but moving the tote stack forward until either the litter contacts the height guide (first stack) or the adjacent stack's RC seems to address the problem (maybe).
