Versions Compared

Old Version 9

changes.mady.by.user Megan Spee

Saved on

compared with

New Version Current

changes.mady.by.user Megan Spee

Saved on

Key

  • This line was added.
  • This line was removed.
  • Formatting was changed.

...

Requested By

Megan Spee

Sub-Team Review

(To be checked once reviewed by sub-team representative)

  •  Mechanical
  •  Electrical
  •  EFS
  •  Autonomy
  •  Operations

Date of Request

2024-03-16

Goal Summary

  • Pegasus Video transmission with different antenna placements, using 5.8GHz

  • OSD configured and MUX functional for front and downwards cameras

  • OFS configured correctly

  • Autolanding on houston Houston functional

Status?

Status
colourYellow
titleWaiting for Sub-Team Review

Desired Airframe

Pegasus
Houston

Location + Time

WrestRC

...

Expand
titleCard 1: Pegasus VTX testing

Video muxing w/ at least 2 cameras. Ideally 3: 1 front, 1 bottom (monster mount), 1 bottom (cabin)

Test #1: Pegasus VTX, LOS low speed & hover performance

Procedure

Goals / Objectives

Failure criteria

  1. Takeoff manually in Loiter, verify that video is stable during takeoff.

  • Ensure that video signal is maintained during close-range takeoffs.

  • Ensure hover doesn’t negatively impact video switching

  • Ensure OSD elements are correct

Loss of Video Transmission

  • Stop drone movement, attempt to re-establish video link.

Loss of Control

  • Land the drone as soon as possible

Loss of Orientation

  • RTL

Battery

  • Land, RTL if possible.

  1. Attempt switching video sources

  1. IF CONFIGURED: attempt switching OSD screens.

  1. Fly slow passes at varying altitudes / distances (pilot discretion) from the ground station

  • Ensure that video signal is maintained during different maneuvers

  • Ensure that different video sources are functional during different passes

  1. Land normally

Expand
titleCard 2: Pegasus Houston pathing

Test #2: Auto Pathing Mission

Procedure

Goals / Objectives

Failure criteria

  1. Connect pegasus Houston to Mission Planner

  • Verify connection

  • Mavlink doesn't connect

  1. Run pathing code

  • Go to pathing repository

  • Activate the venv with source venv/bin/activate

  • Run pathing code with python -m path_2024

  • Failed sending waypoints

  1. Check waypoints

  • Verify waypoints added by pathing code are correct in Mission Planner

  1. Takeoff

  • Verify drone flies to correct waypoints

  • Pitch/Roll/Yaw incorrect

    • land, flip in mission planner, re-start test

  • Drone de-stabilizes.

    • Land, evaluate tune, re-start test

Expand
titleCard 3: Houston auto landing

Test #1: Auto Mission Landing

Procedure

Goals / Objectives

Failure criteria

  1. Connect Houston to Mission Planner

  • Verify Connection

  • Mavlink doesn't connect

  1. Verify auto-landing script is running

  • Check that images are being logged

  • Script isn’t running

    • Start it manually by navigating to the auto landing repository and running python blue_only.py

    • Make note so this can be fixed

  1. Add waypoints in Mission Planner

  • Add a LAND waypoint at the intended landing pad

  • Offset the waypoint by ~5m to test if the auto landing script corrects it

  1. Takeoff

  • Verify drone lands on landing pad

  • Drone misses landing pad

...

Explain what capacity you need, what needs to be mounted, etc.

Mechanical

  •  Pegasus cleared to fly (mechanically)
    •  sensor mount ready
    •  Motors + frame rigid
  •  Houston cleared to fly (mechanically)
    •  sensor mount attached
    •  arms + plates rigid

Electrical

...

Name

Phone # (opt)

Sub-team

Driving (question)

Role

Megan Spee

Director

toyota prius

pilot/organizer

Nathan Green

rav4

pilot/GSO

Amy Hu

Autonomy

auto land

Georgia Vachon Westerlund

Ops/Mech

media

Jane Zeng

Autonomy

Yuchen Lin

GSO/organizer

Neel Patel

...

Note

This section and remaining to be filled out by FTC and Flightline Team

...

Media Ingest: https://uofwaterloo-my.sharepoint.com/:f:/g/personal/uwarg_uwaterloo_ca/EqjTFLI7eSZHil5CgZ13PYMBk9KPlNMFPESkOseTALivzg?e=tapYW3

Tip

  • did not crash

  • auto landing worked, relatively accurate

  • pathing worked as desired

  • able to accomplish all FT objectives

  • VTX signal worked consistently

  • flew DJI for some cool media

  • Able to fly houston entirely using mission planner w/o controller

  • Video switching worked!

Note

  • Camera bottom-facing might have some damage; strange colors in shot

  • Didn’t use Pegasus for the automatic pathing

  • LTE worked a bit

  • Barometer position hold was mid (as always, not fixable)

  • Houston compass not working well, probably just needs calibration

  • Camera noisy

  • Arm switches for houston and pegasus not the same on the pink controller

Warning

  • LTE cut out and didn’t reconnect. Multiple times

  • We forgot crossfire tx and gemini on houston. Had to run houston without a controller, just using mission planner

  • EFS optical flow setup wasn’t functional. Had to change parameters for this on the flight line

Action items

  • Test the OSD with a multimeter to see why it was cutting out.

    • they do have a common ground.

    • so maybe the noise induces

    • ground is not 0. becomes noise in the ground line, so the difference between high and low voltage is negligible.

    • wires too long, so reduce the interference

    • Shorten wires for entire video system. move things around.

    • new mount that incorporates both the front and bottom video cameras

  • LTE system, test in north campus

  • look at forcing it to stay on 1 LTE network and not try switching

  • find LTE map for different towers

  • Secure antenna upright harder (tape or something temporary)

  • Test autolanding on pegasus

  • Test pathing on pegasus

  • check transmitters are on the drones

  • Make sure parameters are ready to fly on after configuration

  • calibrate compass on houston

  • Change arm switch on houston model to match peggy

  • Auto mission

    • integrate auto landing with pathing

    • take all the pathing waypoints from the conops and offset them so they’re at wrestRC

    • make the last waypoint go into auto land mode

  • Talk to mech about barbies or equivalent weight simulations

  • process the signal RPI data