The Good Stuff
Vanny flew and competed both days!
Team adaptability and willingness to pivot
Debriefing and applying feedback in real-time
First time competing at an in-person competition!
Mini integration meetings + strict timelines at competition
Things I’m hearing a lot:
Integration test more
Better needs assessment
Test earlier and more often
Issue: Controlled Authorization
Nearly couldn’t fly at the competition because need an advanced
Never set sub-team or team-wide level requirements
Need a designated conops reviewer (either a chief pilot or someone else)
Resolution: At the start of the term, we go through conops as a team, generate a list of requirements encompassing operational and development needs to compete and assign ownership at that time.
Issue: Flightline
No communication during flight initially, wrong tools as we were talking over phone.
No roles during flight, no ownership over various aspects of flight
No pre-flight briefing
Resolution: Purchase radios for communication at flight line, develop pre-flight briefing procedure to assign ownership during flight.
Issue: Timeline
Aircraft was 3-4 weeks behind schedule, no integration until competition
Waiting for everything to be ready rather than flight testing what we have
What held us behind schedule?
IMU issue + ZP not ready early enough
Last minute switch for PDB
Frame rebuild after crash
Resolution: Move flight testing to happen at a consistent cadence rather than dependant on subsystems being ready, so we flight test even if no changes have occurred. More aggressive and fixed go/no-go timelines that are communicated better across the team.
Issue: Harnessing
We didn’t really think of the wiring in advance, didn’t have cases for boards, being more careful around antennae.
No physical layout of boards on drone.
More connectors in general.
Resolution: Centralized wiring diagram and source of truth for what’s on the aircraft.
Issue: RC Dropout
Fresnel effect led to RC dropout at distance
Resolution: Mount antennae on a pole to remove antennae from ground, long-term work on a directional antenna. Talk to prof/grad student about how to improve comms.
Issue: VTX/VRX System
Resolution: Buy a COTS VTX/VRX System.
Issue: Flight stability
Resolution: More robust PID tuning and testing on ZP3 for this. Flight modes to hold drone in place.
Issue: No voltage check
Resolution: Build battery monitoring system using existing telemetry links, further battery testing.
Issue: ZP couldn’t even fly
Resolution: Don’t have a sufficiently iterative design process, tried to do everything at once instead. Going forward focus on basic and stable flight before going for autopilot.
Issue: EE Slow Bringup
Resolution: Not enough iteration and poor allocation of budget meaning boards just weren’t completed.
Issue: Poor material choice for frame
Designs weren’t finalized which forced things to be changed on mech
Parts weren’t ordered as soon as we knew we needed them.
Didn’t iterate and change design enough.
Resolution: Finalize designs earlier with better requirements capturing, iterate faster and order parts sooner. Make ease of changing design, ease of assembly/disassembly, high tolerance mechanical design requirements.
Issue: Weight budgeting
We never kept or maintained weight budgets at all, a rough ones made, had a lot of downstream issues.
FW CV EE never gave lists of what was going on the drone.
Resolution: More subteam integration discussions and maintaining weight budgets in advance.
Issue: Sleep deprivation
Resolution: Add a slack day, arrive 1 day early to competition always
Issue: CV Software Environment
No documentation
Little integration testing esp since comms was broken
Resolution: Use docker, don’t use QT anymore, top down paydown of technical debt, cv lead owns docs.