On the APM developer mailing list, the team recently posted their software roadmaps. Everyone is welcome to join the dev list themselves and follow this, but for those who aren't already signed up, I thought I'd post them here.
As most of you know, the big recent leap has been full ArduCopter support for the 32-bit PX4 autopilot board, which is now complete and we'll talk about more soon. But in the meantime, here's the ArduCopter software near term to-do list, from team lead Randy Mackay (some of these are already in ArduCopter 2.9.2, which is in beta testing and will be released soon)
- inertial navigation for loiter and waypoints
- compensation for motor interference of compass that leads to heading inaccuracy
- limits (aka geofencing) rework to make more reliable and use less memory
- px4 support
- bug fixes and small enhancements including:
- single channel failsafe detection
- make smoother throttle transition from manual <-> auto flight modes
- rssi pin
- user controlled yaw-behaviour parameter for auto missions
- auto loiter command fix-up (ignoring altitude provided by user)
- auto ignoring altitude as requirements for completing a waypoint
- CH6 doesn't work for tuning when gimbal is enabled
- waypoints don't output in dataflash like on arduplane meaning you can't see the waypoints in goggle earth
- fix flip code
- i2c sonar
- do_jump is broken
- fix hil
- send failsafe message to ground station
Personally I don't plan more than about 1~2 months ahead because priorities change and because team members are mostly free to work on what interest's them so you end up with happy surprises while also having issues that don't get solved for a long time (especially lower priority items).
Comments
Hi Chris,
You said in a recent interview that when the Arducopter release version of the PX4 code was "released" that you (3DR) would be introducing a revised and enclosed version of the PX4.
I am waiting with baited PayPal card.
Just a thought, as you may know I am writing a section in the Old Arducopter Wiki about vibration control.
You might want to think about including adequate for inertial navigation vibration control in the new PX4 enclosure.
Just a thought, somehow, I'll bet one you've already had, but I had to say it.
Absolutely fantastic to hear this.
PX4 and Optical Flow definitely next on my to get list.
In the meantime I've been working on the new anti-vibration section of the Wiki which is going to be really critical for inertial navigation.
Thank you very much developers.
Jason
About that CH6 gimbal thing. If you reverse the gimbal in camera setup it will also reverse CH6 for tuning, so you may not have the setting you think you do.
Along with Ardupilot Mega Planner, some sort of maintenance mode will be essential that would help to determine failure of devices and mechanical mismatch such as vibration. Fast Fourier Transform will help a lot to analyze where the noise comes from.
Chris, Great to hear ! Names of Doug and Jasson are not mentioned as associated with Arducopter. These guys have done great job.
* compensation for motor interference of compass that leads to heading inaccuracy
This one sounds very interesting! I wonder how they do that?