...
Goals for this test
Complete a 1.5 km range test on Pegasus 2 to measure dropped packets
Do this without even flying the drone, put it in a car, drive it a kilometer down the road, measure packet loss.
Flying the drone is a risk
max distance we need at comp is 1.15 km
Can we make sure that when we do range tests, we log continuous data on both tx/rx side? Would be nice to know how our tpwr fluctuates as a f'n of distance
Can we try to test different antenna orientations? I.e. if we rotate the drone does it severely impact signal quality.
Can we try this test with/without other ELRS rx's active (maybe for ex: houston) near the pegasus basestation? I know they say elrs<->elrs interference is low but I want to see how low
Test the analog video system and relay system on Pegasus 2
Asana tasks to set this up
does not require flight, do at same time as mavlinkrc rank test
Test Geolocation system on Houston
Tasks in Asana to set this up
this will require
driving to wrestrc
Integrate and test hardware leading up to this
RPi integration
Autonomy preparation
Geolocation software exists and will be run on Houston
houston is ready for this test hardware wise
Try to isolate altitude, flight pass-by speed, and distance from take-off/home location as variables for our testing geolocation accuracy
Will be important to determine altitude and speed at which we scan for hotspots later in comp
Attendees
If you do not put your name here we will not assume you are coming.
Nathan Green is not coming
Anthony Luo is not coming
Evan Janakievski primary pilot, SDC vehicle driver
Will not be able to drive personal vehicle because driving SDC vehicle
Alison Thompson mech lead + car (fits 4 + driver)
Daniel Puratich supervision & tuning
Ashish Agrahari autonomy
Tochi Okoro autonomy
Maxwell Lou autonomy
...
Start charging batts at 9am
only charge one set of batts for Pegasus 2
Leave Bay 10am
Debrief starts at 12Done 1pm 1
Done 2pm
Debrief
Intro
Attendees:
most everyone else left or wasn’t interested
Objective of debrief: diagnosing and resolving a few difficulties this flight test + the flight test preparation that resulted in not being able to fly today.
General
General topics - very loosely defined by Megan Spee , who was not actually at the flight test (and improved by Daniel Puratich who was)
Not enough preparation for this flight test
Houston configuration issues
houston was working last week but we wanted the rpi killswitch to work
params were changed to try to do this
changing these params broke the build and a motor would not spin
we were not able to resolve at flight line
parameters backups were either
not existent
not properly organized so we could find them
Peggy telemetry setup issues
didn’t know how to get ardupilot to connect to the controller for telemetry
spent 30 mins to get the controller to connect to drone at all
Unable to test auto code or ground test peggy as a result.
will need to redo this test, likely after midterms
Note that auto code (geolocation) was successfully ground tested on houston the day before the flight test (Ashish Agrahari)
systems integration / flight test operator / pilot knowledge base is lacking.
documentation does not exist. difficult to make this exist, because it’s related to:
ELRS (constantly changing)
Ardupilot (documentation already exists online for a lot of this, we don’t want to re-write for obsoletion)
things just change on the drones all the time and people don’t have time to keep up.
Houston and Peggy fly guides are out of date or unfinished.
controller inputs and curves are confusing and the documentation is out of date.
General sentiment of better-clarifying who is responsible for what at each flight test.
Changing around the flight test owners frequently so more people get exposure
Having experienced members walk newer pilots through these steps before being left to do these tasks by yourself.
you don’t know what you don’t know!
Directors should delegate the task management with PMs better to avoid overloading directors
Daniel had a busy week, missed comp arch sync
Couple thoughts on the fixed wing.
Evan (airframe lead) unable to attend fixed wing syncs to express concerns
‘too late’ to fix a lot of design flaws
low confidence it will work well with the current manufacturing thoughts
Action Items
Action items following debrief:
Use the flight test attendee roles harder.
Ping people more leading up to the flight test.
Nothing is too overwhelming of a task for one person if you divide up the work enough.
comp arch sync is a good time to go hard on these conversations, we need to use that time!
All directors, pilots, flight test coordinators should take turns owning a flight test.
the admin overhead to make sure tasks are being done leading up to the FT and deciding what we need to do in prep for flight tests is a lot of effort, needs to be split around
Make sure knowledge transfer is happening.
Have more experienced members coach/work through the setup of peggy and houston before flight tests (ground tests), and at flight tests.
can also be aided by documentation.
Maybe some old operation guides for examples too.
letting people try things with supervision is important before they can do it on their own