Will Autoquad firmware be of any use to us?

This forum is dedicated to software development related to MultiWii.
It is not the right place to submit a setup problem.
Software download
Post Reply
Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Will Autoquad firmware be of any use to us?

Post by Katch »

Some of you may be aware that the Autoquad team have finally finished tinkering and are coming to market with their first set of test products (ESCs and FC).

The FC is an incredibly piece of hardware and software; Full auto takeoff/landing, POI navigation and orbiting, control of speed and alt to waypoints etc etc. If the video is a true representation of everyday performance then they are doing a lot of things right.

http://www.rcgroups.com/forums/showthread.php?t=1683776

The autoquad team are keeping the software open source so will it be any use to the MWC project? Will we be able to port function across and improve our own firmware with their work?

ronco
Posts: 317
Joined: Thu Aug 18, 2011 2:58 pm

Re: Will Autoquad firmware be of any use to us?

Post by ronco »

thay did a lot of things right.. but did you look at the caliration of it?

thay are a lot of steps needed to get this thing to work .. .. freeze it .. heat it up... i think thay started developing it with a complete different point of view as it was as multiwii was created..

thay can optimize for just one board (always the same sensors)
and thay did it without taking care of simpleness..

i belive there are things that we can use but i dont think that we can get a auto flying mode that works like the autoquad's does..

but i also think that this is not the definition for MWC ;)

regards felix

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

So many of our current difficulties in getting good alt hold, solid navigation etc stem from poor calibration techniques. I completely agree their's is a different beast altogether but there are lessons to be learned and techniques to be appropriated into our controller.

timecop
Posts: 1880
Joined: Fri Sep 02, 2011 4:48 pm

Re: Will Autoquad firmware be of any use to us?

Post by timecop »

Not as long as you keep stuck to obsolete 8 bit hardware and wii sensors.

copterrichie
Posts: 2261
Joined: Sat Feb 19, 2011 8:30 pm

Re: Will Autoquad firmware be of any use to us?

Post by copterrichie »

dongs wrote:Not as long as you keep stuck to obsolete 8 bit hardware and wii sensors.


But we are talking apples and oranges here. The Wiicopter started as an accessible to everyone hobby but I have to agree, it has grown to much more.

User avatar
EOSBandi
Posts: 802
Joined: Sun Jun 19, 2011 11:32 am
Location: Budapest, Hungary
Contact:

Re: Will Autoquad firmware be of any use to us?

Post by EOSBandi »

DNFTT

timecop
Posts: 1880
Joined: Fri Sep 02, 2011 4:48 pm

Re: Will Autoquad firmware be of any use to us?

Post by timecop »

EOSBandi wrote:DNFTT

dude, seriously? OK, knock yourself out writing 16-state ukf on avr. maybe you'll run out of flash on mega and be able to run it at 1hz.
If you don't have anything constructive to say to my comments... don't say anything?

jy0933
Posts: 180
Joined: Wed Jun 27, 2012 4:24 pm

Re: Will Autoquad firmware be of any use to us?

Post by jy0933 »

my understanding of their initial goal is pc controlled drone... more on point to point movement.. kind of patrol like

mwc atm is manually control at most..

also.. I think it wont be hard to improve mwc to AQ stage.. after PID set

altitude control : auto-level w/ baro & sonar( rpm feedback from esc might do better since you can just lock down the rpm)
waypoint: mag+gps ( we already can do gps home .. waypoint is just an extension. the pathlog can be obtained from sending gps cood to ground station)
taking off/ landing control: z axis accelerometer will contribute to measure acceleration on different throttle( kind of calibration and also decide the taking off/landing speed.. ).. an addition is sonar sensor to detect close ground distance and velocity
speed control: I have no idea.. but I do see ppl doing fvp having ground speed on the screen.. some1 name that sensor for me? (thx)
telemetry: shoot all the data to ground station>> PC
better tx/rx system: shoot pc control sig to quad


what else?!
A great programmer(team)

problems?
optimization for all compatible board will take a freaking long time .. while choosing one to start with might seem unfair to the other board users...
cost of integration of all those components/sensors... single cost is not high/mass prod cost is not either.. but a integrated board means redesign of a pcb.. or a lot of wires going around :o

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

MultiWii started as a sports multirotor controller for sure and that is still one of the things it does best but Alex and the devs are taking it far beyond that with the goal of being a fully featured controller with GPS and autonomous functions whilst remaining scalable and affordable.

GPS is finally working nicely and being integrated into a stable release and further functions will undoubtedly follow (Auto takeoff and landing) - POI orbit, waypoint navigation not to mention the integration of things like Sonar and Optical sensors and a more robust serial link for telemetry.

I would imagine there are quite a few useful things in the Autoquad firmware that could be adapted and ported to MWC.

PS - I'm going to need someone to lock me away with no internet access and no credit cards tomorrow when the Autoquad FC goes on sale - I will not be an early adopter... gah but I want to be.

User avatar
EOSBandi
Posts: 802
Joined: Sun Jun 19, 2011 11:32 am
Location: Budapest, Hungary
Contact:

Re: Will Autoquad firmware be of any use to us?

Post by EOSBandi »

Katch wrote:MultiWii started as a sports multirotor controller for sure and that is still one of the things it does best but Alex and the devs are taking it far beyond that with the goal of being a fully featured controller with GPS and autonomous functions whilst remaining scalable and affordable.

GPS is finally working nicely and being integrated into a stable release and further functions will undoubtedly follow (Auto takeoff and landing) - POI orbit, waypoint navigation not to mention the integration of things like Sonar and Optical sensors and a more robust serial link for telemetry.


And, all of these functions will work on the old outdated 8 bit architecture, all what need is more memory. So expect that future nav functions will be 1280/2560 only.
Mr.Dongs will port these future functions for new contemporary 32bit whizzbang architecture, but porting to a better processor will not autonomically improve these functions.

Pyrofer
Posts: 180
Joined: Sat Apr 14, 2012 2:55 pm

Re: Will Autoquad firmware be of any use to us?

Post by Pyrofer »

And in anticipation of this I ordered a 2560 flight control board.
Looking forward to trying GPS features, especially programmed routes.

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

It's nice that several very affordable options for 2560 FCs have recently hit the market. I wonder how much longer the 328 can hold on. I guess it will always have a place as a sport flying controller but I think it's days are numbered in development terms.

I wonder if it might almost be time to wrap up 328 as legacy hardware. MWC flies about as well as we could ever hope in Level and Acro but moving forward we need the extra memory that 2560 and STM32 based boards bring.

jy0933
Posts: 180
Joined: Wed Jun 27, 2012 4:24 pm

Re: Will Autoquad firmware be of any use to us?

Post by jy0933 »

how abt naza 32? board.. I didnt check the memory .. but faster processor...and pretty good price

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

jy0933 wrote:how abt naza 32? board.. I didnt check the memory .. but faster processor...and pretty good price


Naze32 is already part of the MWC family running Dongs' baseflight which is a port of MultiWii and uses an STM32 processor.

There's plenty of talk and reason to move to a new primary platform - it's more a question of where and when we jump not IF. 2560 will tidy us over for maybe another 12 months but it's not the ultimate answer. I think there may be a hope that Arduino's own 32bit platform (Due) will be the next primary platform but I'm not so sure.

STM32 seems to be the logical choice and one that has already been made but other FCs in the market but this jump brings with it a whole host of challenges to our dev team that are pretty entrenched in Arduino IDE (this is in no way meant as a criticism)

Maybe we are getting to the point where we need to make some decisions about 1) what MultiWii will be in the future and 2) what architecture best suits that purpose.

The answer to 1) seems to be self evident in the character of current development. MWC is aiming to be a fully featured controller with autonomous capability. For what exactly this will include we need to look at what MK, Autoquad and to a lesser extent APM/MegapirateNG are doing. I also see MWC as being a highly experimental platform that encourages people to try new things and ways to work with multirotor aircraft. I hope this continues.

The answer to 2) is the real issue these days but it is definitely not the 328 and the 2560, as I have said already, is more of a stop gap than a realistic choice moving forwards.

copterrichie
Posts: 2261
Joined: Sat Feb 19, 2011 8:30 pm

Re: Will Autoquad firmware be of any use to us?

Post by copterrichie »

I think the fundamental question that needs to be addressed is, Where is the line where the WiiCopter stop being a hobby and becomes a commercial application? Or can it support both worlds?

Pyrofer
Posts: 180
Joined: Sat Apr 14, 2012 2:55 pm

Re: Will Autoquad firmware be of any use to us?

Post by Pyrofer »

It will never be commercial as it's open source.
People can sell hardware kits, but MultiWii fills a VERY important niche, its free, its easy and its hardware requirements are simple and low.

Going forward there are many options, but I think its important that those roots are not forgotten.

That said, feature tradeoff will very quickly make a 328 look like a bad option. I know that the 328 will always be fine for great basic flight. Its the advance stuff that will need bigger hardware, which is fair enough.

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

I like to think of us as the Linux of the Multirotor world. I hope we can continue doing what we do in the way we do it without moving to closed hardware and commercially motivated development.

I don't know what provisions and deals are made with people that commercialize MWC based hardware but I MultiWii would not work as a completely commercial endeavour - it would cease to be what it is and become another MK or DJI or Autoquad.

I think we attract some very talented coders and contributors by the very nature of the project being open source hardware and software. The moment you commercialize you have a private party of developers and each needs to be sufficiently rewarded for their contribution.

jy0933
Posts: 180
Joined: Wed Jun 27, 2012 4:24 pm

Re: Will Autoquad firmware be of any use to us?

Post by jy0933 »

i think it is definitely moving forward... there is no reason staying not to jump to new faster chip platform to gain either better performance and opportunity(more functions possible ) while maintain at a relative low/reasonable price.


my answer to 2. the more functions add to fc the better proc is needed. liek you said..mwc is successfully ported to stm32 and working pretty decent... ard32 is good to heritage the current platform w/o much mod needed ( I guess... ) either of them will be a good choice.. 328/2568 might be a good step stone.. but the trend, as what i can see, is heading toward 32bit.. (thanks to tc and AQ) >> and why.. as soon as they succeeded, MASSIVE copycats might come up in cheap. but the copycats usually dont have good fw/sw... that would be where mwc coming in the game.( if compatible, probably will.. so the copycats can sell faster and more)

( unless all extra functions are linked back to ground station to process. that means fc can just maintain the most basic functions)

copterrichie
Posts: 2261
Joined: Sat Feb 19, 2011 8:30 pm

Re: Will Autoquad firmware be of any use to us?

Post by copterrichie »

Linux is a perfect example, many MANY corporations are using Linux in their data centers and many companies sell their servers with Linux installed, IBM happens to come to mind. Clarity, when I said Commercial, I was implying the same as the road Linux has taken. Where is the problem of having a Police Drone using WiiCopter Firmware?

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

Pyrofer wrote:
That said, feature tradeoff will very quickly make a 328 look like a bad option. I know that the 328 will always be fine for great basic flight. Its the advance stuff that will need bigger hardware, which is fair enough.


Which brings me back to an earlier comment - maybe it is time to tie up loose ends on 328 and then put it to bed. Anyone with a 328 can now have excellent Level and Acro flight, acceptable Alt hold and basic GPS functions (either via i2c GPS or swapping the serial port over after config). The 328 will not really be able to ever deliver much more than this.

It would make sense to have a final version for 328 that has no further features added and only bug fixes until such time as no more bugs are found.

As we are open source this will not stop people adding new features to 328 platforms (or porting over new features from the main branch) but all this can be done aside from the main development in forks.

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

copterrichie wrote:Linux is a perfect example, many MANY corporations are using Linux in their data centers and many companies sell their servers with Linux installed, IBM happens to come to mind. Clarity, when I said Commercial, I was implying the same as the road Linux has taken. Where is the problem of having a Police Drone using WiiCopter Firmware?


And this is exactly what Flyduino, Multiwiicopter.com and many others do. They sell hardware based on our GPL software. Many people that produce this hardware and commercialize MWC code also modify it and in turn contribute those modifications to main codebase. This is the exact model that Linux follows with commercial enterprises such as Canonical (Ubuntu) developing and providing support for the opensource software.

But the core devs and contributors, us, are a separate layer to these commercial enterprises and I believe we will, and should, remain so. Mutliwii.com will always be the heart of this open source project and I do not see it becoming a commercial entity in its own right.

jy0933
Posts: 180
Joined: Wed Jun 27, 2012 4:24 pm

Re: Will Autoquad firmware be of any use to us?

Post by jy0933 »

speaking of commercial... this might sound interesting... board manufacturing & retailing. preloaded with mw then donate portion of profit to the community

(this is what I saw from reflash esc in RCG... that is not a massive prod yet.. but could be a good model...also I wonder the if commercial projs such as flyduino does donate for the community aside from codebase? just curious)

copterrichie
Posts: 2261
Joined: Sat Feb 19, 2011 8:30 pm

Re: Will Autoquad firmware be of any use to us?

Post by copterrichie »

I think it would be a great idea, Maintaining this website is not free I am sure. :)

jy0933
Posts: 180
Joined: Wed Jun 27, 2012 4:24 pm

Re: Will Autoquad firmware be of any use to us?

Post by jy0933 »

copterrichie wrote:I think it would be a great idea, Maintaining this website is not free I am sure. :)


yeah.. maintaining the entire community is not a cheap thing.. so does the dev kits.

but when it is direct related to the community/proj.. need a good way to prevent people from thinking it is mwc community doing commercial.. and also. several commercial mwc board has already taken a big portion of the users.. how to cut in the market... (at least need to start getting profit then got money to donate) would be a problem.. (mwc with gyro/acc/baro,mag is only 54.. hard to beat the price)

User avatar
kos
Posts: 286
Joined: Thu Feb 16, 2012 4:51 am
Location: Fr

Re: Will Autoquad firmware be of any use to us?

Post by kos »

dongs wrote:
EOSBandi wrote:DNFTT

dude, seriously? OK, knock yourself out writing 16-state ukf on avr. maybe you'll run out of flash on mega and be able to run it at 1hz.
If you don't have anything constructive to say to my comments... don't say anything?


http://en.wikipedia.org/wiki/Fallacy

In logic and rhetoric, a fallacy is usually an improper argumentation in reasoning often resulting in a misconception or presumption. Literally, a fallacy is "an error in reasoning that renders an argument logically invalid".[1] By accident or design, fallacies may exploit emotional triggers in the listener or participant (appeal to emotion), or take advantage of social relationships between people (e.g. argument from authority). Fallacious arguments are often structured using rhetorical patterns that obscure any logical argument.

User avatar
kos
Posts: 286
Joined: Thu Feb 16, 2012 4:51 am
Location: Fr

Re: Will Autoquad firmware be of any use to us?

Post by kos »

Katch wrote:The autoquad team are keeping the software open source so will it be any use to the MWC project? Will we be able to port function across and improve our own firmware with their work?

short answer , yes

User avatar
EOSBandi
Posts: 802
Joined: Sun Jun 19, 2011 11:32 am
Location: Budapest, Hungary
Contact:

Re: Will Autoquad firmware be of any use to us?

Post by EOSBandi »

kos wrote:
dongs wrote:
EOSBandi wrote:DNFTT

dude, seriously? OK, knock yourself out writing 16-state ukf on avr. maybe you'll run out of flash on mega and be able to run it at 1hz.
If you don't have anything constructive to say to my comments... don't say anything?


http://en.wikipedia.org/wiki/Fallacy

In logic and rhetoric, a fallacy is usually an improper argumentation in reasoning often resulting in a misconception or presumption. Literally, a fallacy is "an error in reasoning that renders an argument logically invalid".[1] By accident or design, fallacies may exploit emotional triggers in the listener or participant (appeal to emotion), or take advantage of social relationships between people (e.g. argument from authority). Fallacious arguments are often structured using rhetorical patterns that obscure any logical argument.


:D

I browsed through autoquad code and found some intersting stuff, but nothing extraordinary. The only thing that makes me think is the gyro and acc temperature compensation. It's currently ignored in MultiWii code, given the assumption that we will recalibrate at every battery change. However the we don't care about the changing bias. Perhaps that part would be interesing. (BUt it obviously not needed to freeze the board then warm up, 5-6degree C is fine for calculating...

Katch
Posts: 280
Joined: Thu Aug 04, 2011 1:44 pm

Re: Will Autoquad firmware be of any use to us?

Post by Katch »

I'd love to see Auto takeoff/land and POI orbit making its way into our GPS functions

Alexinparis
Posts: 1630
Joined: Wed Jan 19, 2011 9:07 pm

Re: Will Autoquad firmware be of any use to us?

Post by Alexinparis »

I would globally say yes.
This is one of the strength of open source projects. There is nothing to fear to take a good idea if we can use it.
The last example is the GPS code taken originally from arducopter project, adapted and customized to multiwii.

Post Reply