Working on a new VTOL platform that isnât considered as much. VTOL ground rover instead of a VTOL plane. VTOL tiltrotor ground rover to be exact. Since groundrover is really just a fork of the FW controller with some extras on top. Does any one think that by modifying the current VTOL attitude controller and tiltrotor control, one could just swap âgenerally speakingâ any FW instances with groundrover, and remove any specific transition code relating to in flight transitioning and be done with it?
My first thought is that youâd want ALL the controllers though (GND, MC, FW). Full ground rover support, vertical takeoff, transition to forward flight, and full FW support.
Sweet! Iâll need to modify the ground control a bit for skid steer and possibly reverse if that hasnât been fully implemented in autonomous or offboard modes yet and if my escâs are even capable of reverse to begin with. But other than that, Whyâd you suggest integrating the FW controller too? Oh, this is strictly a multicopter in the air. Integrating wings into the design is another whole realm of complexity in design I donât have time for or what project requirements call for. That would be a pretty sweet future project though.
I may have imagined something different. Could you describe the full vehicle? I was thinking of a vtol vehicle (eg quadplane) + rover, but are you saying itâs more of a rover + quadcopter?
Either way the best place to start is to fill out the rover functionality thatâs missing (reverse, etc). After that weâll need a small amount of logic to switch to flight mode, but I can help with that.
Ah, it is a rover + quadcopter. I said tiltrotor before because I was thinking of having the rotor housing be the wheels. For starters, yeah adding that rover functionality in, and working towards just integrating a simple ground/quadrotor control scheme switch is enough for the concept stage.
Ok, that should be relatively straightforward, itâs just not going to work out of the box without a small amount of Firmware side work.
Iâd start with getting a good rover config working, then weâll make it possible to run the multicopter controllers simultaneously without fighting for the same outputs. Then we expand the transition commands so that you can switch back and forth between Rover <-> Multicopter, rather than MC <-> FW.
Small amount of firmware works sounds reinforcing of my initial idea of not having to reinvent an entirely new controller but one that contorts existing ones together.
Okay, cool. Just need to develop the physical vehicle. Although simulations could suffice for the meantime to work out the ground-rover extensions and controller switching. Iâm thinking a simple H frame quadcopter with wheels on the outside of the frame and props on the tip oriented up like a typical quad. Tilting could come in later stages.
