Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Jump to bottom

Unable to boot on Raspberry Pi 2 #223

Closed
joevo2 opened this issue Jan 5, 2017 · 24 comments
Closed

Unable to boot on Raspberry Pi 2 #223

joevo2 opened this issue Jan 5, 2017 · 24 comments

Comments

@joevo2
Copy link

joevo2 commented Jan 5, 2017

Is there a hard dependency on RPI3? I tried to boot the latest release of Farmbot OS on RPI2 but it wouldnt work I only get the rainbow boot screen
@ConnorRigby
Copy link
Contributor

It shouldn't be to hard to make a config for rpi2, but Farmbot's configuration relies on a wireless interface, which previous generations don't have. If you have a usb wireless dongle, i can probably set something up

@joevo2
Copy link
Author

joevo2 commented Jan 6, 2017 via email

@ConnorRigby
Copy link
Contributor

yeah you would be able to, but you would have to access farmbot configuration by your farmbot's ip address. Not impossible, you will just have to find your ip address information. i will look into adding support for RPI1 and RPI2 soon

@joevo2
Copy link
Author

joevo2 commented Jan 6, 2017 via email

@ConnorRigby
Copy link
Contributor

im starting work on it right now. I don't own an RPI2 any longer, so i may need some help debugging.

@joevo2
Copy link
Author

joevo2 commented Jan 6, 2017 via email

@joevo2
Copy link
Author

joevo2 commented Jan 8, 2017 via email

@ConnorRigby
Copy link
Contributor

if you have a rather beefy linux machine you can build this:
see BUILDING.md for instructions. Please be warned that this is probably pretty unstable at the moment.

@menghong777
Copy link

Hi, @ConnorRigby . I was trying to build the RPi2 image on Ubuntu 16.10. I got an error when I tried to run "mix firmware", it returns "** (Mix) Could not locate System". Do you have any idea with this error? =)

using rpi2 configuration.
warning: variable "deps" does not exist and is being expanded to "deps()", please use parentheses to remove the ambiguity or change the variable name
  /home/menghong/farmbot-image/farmbot_os/deps/rpi2/nerves_lib/mix.exs:11

warning: variable "deps" does not exist and is being expanded to "deps()", please use parentheses to remove the ambiguity or change the variable name
  /home/menghong/farmbot-image/farmbot_os/deps/rpi2/nerves_lib/mix.exs:14

warning: variable "deps" does not exist and is being expanded to "deps()", please use parentheses to remove the ambiguity or change the variable name
  /home/menghong/farmbot-image/farmbot_os/deps/rpi2/amnesia/mix.exs:7

warning: variable "package" does not exist and is being expanded to "package()", please use parentheses to remove the ambiguity or change the variable name
  /home/menghong/farmbot-image/farmbot_os/deps/rpi2/amnesia/mix.exs:8

warning: variable "docs" does not exist and is being expanded to "docs()", please use parentheses to remove the ambiguity or change the variable name
  /home/menghong/farmbot-image/farmbot_os/deps/rpi2/nerves_firmware/mix.exs:17

** (Mix) Could not locate System

@ConnorRigby
Copy link
Contributor

sorry about that, the system needs to be built first. for clarity when we say system it means the root filesystem built by buildroot that boots linux.
so to build the system cd to the root of the farmbot_os dir, then do this:

export MIX_ENV=prod
export NERVES_TARGET=rpi2
mix farmbot.system # this builds teh environment for building the system.
cd apps/NERVES_SYSTEM_${NERVES_TARGET}
make

that is going to take a while depending on your machine. once it is finished do this:

cd ../farmbot
mix firmware

@menghong777
Copy link

Thanks for your reply and instructions. =) Before running the make command, I followed Nerves Project website's instructions (https://hexdocs.pm/nerves/systems.html#creating-or-modifying-a-nerves-system-with-buildroot) then I ran mix firmware command. After that I got an error shown below.

==> nerves_network_interface
cc -c -I/usr/lib/erlang/usr/include -O2 -Wall -Wextra -Wno-unused-parameter -pedantic -std=c99 -D_XOPEN_SOURCE=600 -o src/netif.o src/netif.c
src/netif.c:26:27: fatal error: libmnl/libmnl.h: No such file or directory
 #include <libmnl/libmnl.h>
                           ^
compilation terminated.
Makefile:51: recipe for target 'src/netif.o' failed
make: *** [src/netif.o] Error 1
could not compile dependency :nerves_network_interface, "mix compile" failed. You can recompile this dependency with "mix deps.compile nerves_network_interface", update it with "mix deps.update nerves_network_interface" or clean it with "mix deps.clean nerves_network_interface"
==> farmbot
** (Mix) Could not compile with "make" (exit status: 2).
Depending on your OS, make sure to follow these instructions:

  * Mac OS X: You need to have gcc and make installed. Try running the
    commands "gcc --version" and / or "make --version". If these programs
    are not installed, you will be prompted to install them.

  * Linux: You need to have gcc and make installed. If you are using
    Ubuntu or any other Debian-based system, install the packages
    "build-essential". Also install "erlang-dev" package if not
    included in your Erlang/OTP version. If you're on Fedora, run
    "dnf group install 'Development Tools'".

I'm sure I have installed gcc, make, build-essential and erlang-dev. This is my first time compiling an operating system. Am I doing it right? =)

@RickCarlino
Copy link
Contributor

@menghong777 You will need to install libmnl, also.

@ConnorRigby
Copy link
Contributor

@menghong777 @joevo2 You'll both be happy to know Raspberry Pi2, Pi1, and Pi0 support is now stable. Next release will include support for it, though you will still need to build it yourself. We are working hard to make sure our building documentation is ready for new comers also. Check back soon!

@joevo2
Copy link
Author

joevo2 commented Jan 11, 2017 via email

@ConnorRigby
Copy link
Contributor

@joevo2 i don't have one to test, but FarmbotOS itself only uses about 18 megs of ram and 10 megs of disk space. (the disk space one is subjective depending on how you look at it).
After linux passes control to Erlang all the cpus (4 on rpi3, 1 otherwise) sit idle at about 0.93 mhz, and consume about 4.77 volts. FBOS is pretty lightweight for something considered an OperatingSystem

@menghong777
Copy link

Thanks everyone! Really appreciate everyone efforts in this project and to helping me out! I'm building it using the latest source code now. Hope the compile process is running smooth as expected! =D

@joevo2
Copy link
Author

joevo2 commented Jan 11, 2017 via email

@ConnorRigby
Copy link
Contributor

@joevo2 we are actually starting work on that right now. We plan to have it happen on the rpi i believe. The current implementation happens in python right now.

@joevo2
Copy link
Author

joevo2 commented Jan 11, 2017 via email

@roryaronson
Copy link
Member

We have a prototype made that uses Python and OpenCV to process the images to detect all plants. That code could run anywhere (laptop, pi, in the cloud, etc) though we're planning on having it run on the pi. The code currently outputs coordinates and sizes of all plants (both weeds and our desired crops). This list will then need to be cross-listed with the list of current crops so that we can categorize what is a weed and what is our veggies. Then the user will be able to schedule sequences to run on the weeds to remove them.

Here is a mockup of what it will look like on the web app for the user to "tune" the processing algorithms. It is a two step process: first is selecting the range of colors to isolate and look at, and second is the openCV stuff that locates the individual plant's centers within the image.
weed detection settings widget

@joevo2
Copy link
Author

joevo2 commented Jan 11, 2017

Wow awesome looking! We were looking into this code of yours https://github.com/FarmBot-Labs/plant-detection to modify it for the tray seeding function. It seems like there will be quite a lot of work for us to customise and learn to build a tray seeding function 😢

@joevo2
Copy link
Author

joevo2 commented Jan 11, 2017

Any tips on where we could start with building the tray seeding function? Since it is pretty similar with the weeding functionality. Any code for us to reference? Sorry for the trouble again and super thanks for the hard work being done. We would love to help out. Btw me and @menghong777 is actually on the same team in case you wonder :)

@roryaronson
Copy link
Member

The easiest way for you to accomplish tray seeding is to make a hardware jig that allows you to position your tray precisely in the FarmBot workspace. Then you could just use the regular sequence builder to load the various tray sockets with seeds.

If you're set on a camera based solution then you'll need to identify the z-height, X/Y position, and angle of the tray. You would do this by having easily identified markers on the corners of the tray. Then your code would output socket locations (coordinates) which your "seed loading" sequence would need to use.

We were just discussing today how we plan to allow not only "Plant", "Weed", and "Tool" objects in the database, but also any other type of object with coordinates and a label. This is a little far off I think, but when such a feature is developed your code would be able to create "Socket" objects in the database which then your Sequences could import coordinates from. Hope that makes sense. This is basically how weed detection puts the weeds in the db.

@ConnorRigby
Copy link
Contributor

@joevo2 @menghong777 Starting with release 2.1.9 we added support for other systems. You should download the corresponding .img file and flash it to your SD card.

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment
Assignees
No one assigned
Labels
None yet
Projects
None yet
Milestone
No milestone
Development

No branches or pull requests
5 participants
@joevo2 @RickCarlino @roryaronson @menghong777 @ConnorRigby