Home        Store        Docs        Blog

QGC support for an AUV frame


What an interesting group. Hello All. I’m betting that somewhere in this lot are those who could help me in one way or another with my own peculiar build.

I’m determined to create a uniquely minimalistic AUV architecture. It’s already been a long-term project, owing to its ambitiousness, but it continues to inspire me to learn. I’m coming at it as a seasoned optomechanical engineer with rusty coding skills, still new to Linux, the Open-Source world and its ways.

I’d like to use QGC-BR for all the great work that’s gone into its subsea functionality, but what I’m building doesn’t map to any of the ROV frames on offer. My AUV frame is best described as a twin-motor delta wing “plane”, i.e., with elevons and differential thrust. I’m using the BeagleBone Blue (with ArduPlane) for my FC and I/O.

How can I bridge that gap between QGC-BR and this “airframe”? At this point I don’t understand the magnitude of that effort, or how to best tackle it. Can someone point me in the right direction? Thanks.

(Patrick José Pereira) #2


If you are already using ArduPlane, you should check the original QGC project. QGC BR contains some modifications for ArduSub firmware and share the majority of the features with all vehicles.

(Jacob) #3

We work with the ArduSub firmware in particular on these forums. ArduSub does not currently support forward flying vehicles, but it could be made to work some day. If you choose to use ArduPlane, you will get better help at discuss.ardupilot.org.


Thanks Jacob. It was a coin toss as to which group would be the best starting point for cross-breeding my platform. When you say “could be made to work someday”, can you put that in rough terms of programming man-hours? Weeks? Months? At some point I plan to solicit a team to help out.


Thanks Patrick. I’ll spend some time with the original QGC link over the weekend to get some perspective on this beast. Maybe at some point I’ll be able to help develop an AUV frame family for ArduPilot.

(Jacob) #6

When you say “could be made to work someday”, can you put that in rough terms of programming man-hours? Weeks? Months?

Weeks for initial hacking and proof of concept, months for complete/adequate integration into the codebase.


Excellent. My expectations were in-line with that reality. That programming effort is still some distance away from critical path of the development project, but I’ll need to get organized sometime in the next few months. Thanks again.