Autonomy 2023-2024 Goals & Milestones
Comp Projects
Imacs
QR scanning + GCS Integration
Jetson / Auto-Landing → Airside System
Object detection/classification
Non-Comp Projects
CUCAI
LIDAR(s)
QR Code Scanning + GCS Integration:
end goal
Have a QR Code scanning module that can be run air or ground-side and interface with all of WARG’s Autonomy Repo’s, integrate with GCS Solution to send mavlink messages to create autonomous missions
Priorities
QR Scanner (complete, untested)
DroneKit to interface with mission planner (goal for done by July 16th)
integration painers
Works on ground station laptop
Progress not being made?
nobody on-site to test on WARG Hardware
getting people familiar with dronekit + mission planner
no well defined plan/interfaces for project
Milestones / Check-INs
Goals:
Mid-july:
Dronekit interface works
QR Scanner tested
Test in SITL (goal for done by July 16th)
Early Aug:
Flight test?
EOT S23 this works.
Mavlink Integration: GCS 4 comp
able to interface with OTS or custom GCS & send mavlink messages
end goal
be able to setup AUTO missions using Mission Planner or custom GCS software. (Task #1 or #2 this term)
Priorities
Testing on hardware
Finding edge cases (like geofence issue @ comp)
integration painers
need airframe that can carry large px w/gps (waypoint capable)
w/telemetry
need ground station computer
Milestones / Check-INs
2 months : communicate w/the drone over telemetry & GCS using python script
EOT (4 months): simple auto missions using script (uploading next waypoints, etc)
generating mission files & uploading / running mission files.
November: start to finish auto mission w/ability to restart mission after pilot takeover (see task #2 comp debrief).
This has been integrated into the QR code + GCS stuff
Jetson: Working Jetson System
Jetson working & supported & documented
Xierumeng documentation popoff
Able to communicate with competition system
end goal
Landing pad model integrated successfully onto the Jetson.
Landing pad detection worked on July 9th flight test
Jetson able to communicate over MAVLink or Telemetry link
DroneKit-python api confluence documentation
The DroneKit API provides a high-level interface for communication with drones using the MAVLink protocol, which is a lightweight messaging protocol specifically designed for communication between ground control stations and autonomous vehicles. The API abstracts away the low-level details of MAVLink communication and provides an intuitive set of commands and functions to interact with the drone.
Priorities
Integrating dronekit into cv system
Communication over telemetry
Testing geolocation module
Testing navigation to coordinates
Testing gaussian mixture module? → detect which part of ground has or does not have landing pad.
integration painers
Jetson w/CV Camera
Jetson w/Comp ZP (drivers code complete but untested?)
Test Airframe capable of carrying payload
Milestones / Check-INs
CV Camera + Jetson working
Jetson + ZP3 Working
Camera + Jetson on testing drone.
Goals:
Modules done by EOT.
Yolo Object Detection:
working / supported / annotated ability to collect data, train data, and be deployed on live datasets
end goal
see Jetson
Priorities
integration painers
Milestones / Check-INs
Imacs Revamp : Not Critical For Competition
end goal
Have tested GUI
not possible to create GUI that replaces ardupilot. needs to run alongside ardupilot.
This is more of a task for newer members and doesn’t seem as useful as the others
Priorities
Figuring out data
Test
Work on GUI
Â
integration painers
working device that sends mavlink packets. (simulator or real hardware).
Milestones / Check-INs
Goals:
2mo 4mo: e2e prototype (small, but needs to be e2e)
get info from simulator, and show on GUI
Â
CUCAI
end goal
Participate in the conference and present what we have been working on
Priorities
Look out for registration and plan how we want to present the stuff we have been working on for competition
making a presentation and a booth
showcase what warg has been doing
yolov8 model we made
integration into jetson
progress on extrapolating that into autolanding
integration painers
Big question is how much effort is this to participate in it
can use this for outreach in other places too?
Milestones / Check-INs
2mo: None
applications haven’t openned up yet as of early july
4mo: Look out for conference registration
6mo: Have registration completed and plan our presentation
Lidars
end goal
Â
Priorities
Â
integration painers
It’s either pick something for autonomy to dev with or pick something that works in ardupilot, there aren’t options that do both. we need to look into this, more concerned with other projects.
Â
Milestones / Check-INs
Â