Erle-brain, an open hardware Linux autopilot


Hi everyone,

Some of you might have heard about the work we did with BeaglePilot porting APM to Linux both in the hardware and software side. Work we presented at LibreCon 2014 last month. 

I am happy today to announce that after several months of improvements, flight tests and pre-series with manufacturers we are launching a commercial Linux hardware autopilot based on this work: Erle-brain.

Erle-brain is sold at 269 € and puts together a BeagleBone Black (rev. C) and the PixHawk Fire Cape in a single package that weights about 110 grams and includes 25+ sensors. The hardware designs are open to anyone that wishes to improve them.


The autopilot has a 4 GB eMMC flash memory that comes pre-flashed and provides:

  • Linux 3.8 kernel compiled with the PREEMPT option (best results we measured)
  • Debian Wheezy file system
  • ROS Hydromedusa
  • mavros ROS package
  • APM running natively in Linux (and linked with ROS through mavros)
  • preconfigured daemons for launching everything automatically, WiFi dongles support


Erle-brain has been successfully tested in copters, planes and rovers. Thanks to the contribution of many there're drivers for most of the sensor and we keep working hard to provide support for even more accessories. Here are some of the ones we've been playing with:


Expect more to come :).

Besides doing some hardware hacking we've also been putting time in documenting everything. The APM wiki is great and we love it but we wanted to do it our way so we've spent quite a bit of time creating GitBooks that should provide a walkthrough no matter which is your technical level:



We expect to come up with more material in the next months. Thanks everyone for your support and contributions. We will keep working hard to create amazing Linux autopilots.

Best regards,


E-mail me when people leave their comments –

You need to be a member of diydrones to add comments!

Join diydrones


  • Hey @Shyam,

    It was a pleasure to meet you. Good luck with your projects!

  • @Victor Mayoral, was great to meet and discuss with you at GoToCon yesterday. I remember visiting this page when you posted in late 2014, but overlooked some great information!!

    Thanks for enlightening me, now I am gonna look more closely into Erle-Brain:)

  • Hi Todd,

    You'll need to compile the code with the right setup (change this line) to make use of the magnetometer in the MPU9250. Feel free to reach us at our forum. We'll be happy to help.


    APM Plane, APM Copter, APM Rover source. Contribute to erlerobot/ardupilot development by creating an account on GitHub.
  • @Victor, just got an erle.  while setting up I noticed I must have an external mag for calibration.  Is this typical, or have I missed something?  I haven't been able to keep up with software dev for some time, so playing catch up.  I noticed sensor suit on PXF provides for mag.  Is it just that the current software is not yet making use of the MPU9250? Can you point me in right direction?  Thanks!

  • Philip thanks for your supportive message. We are excited about what is coming with this technology. A few notes:

    • @Damouav, the PXF is available through our store.
    • A ready to fly solution is also available through Erle-Brain.
    • The power connector of Erle-Brain and the PXF now have support for voltage measuring. Refer to this branch.

    We keep working to add more and more capabilities.


  • Developer
    I am working on a 1.6 version of the hardware, it is all working really well, so I will be pushing for a new version soon!
  • Developer

    here is a list of Approved hardware vendors that can use the PXF Trademark
  • Developer

    Hi All

    As the Lead hardware designer of the PXF, I would like to remind you all as to the beginnings and the purpose of the PXF.

    The PXF project came about when I was talking to Andrew Tridgell about his port of APM to the Raspberry Pi as a coding exercise.  Along with some others, Tridge was using a linux co-computer to do search tasks with the CUAV group for the outback challenge. We discussed some of the benefits, as well as some of the concerns of running a full autopilot on Linux, and I won't go into the details other than to say, the system MUST be designed to give the Autopilot absolute priority over every other process.

    So, Preempt, Preempt RT etc etc.... thats an area that Tridge, Victor, Siddharth, and many others have put a heap of work into getting right.

    This is an open project, if you go to you will find the current drawings ( I am working on version 1.6 at the moment)  any Board that Bears the name PXF™ must have MY approval.  The EARL ROBOTICS board has that approval.

    You can look at the HAL code here...

    You will also find a link to a Gitter chat, feel free to jump in, join the effort!

    Why BBB?

       Its OPEN! Its well supported, its well used by many people in robotics research world wide, its capable of running a webserver for direct interaction with the system in a completely device independent manner, while writing code, you can test and compile, on the drone, out at the test site!

        Is it the "BEST" platform for the future? no!  but is it a great platform for now? YES! 

        This project was started as a test effort to guide us in the development of the future of UAV Autopilot hardware, it has become a capable autopilot, thanks to the effort of people like Victor.

    If the BBB changes from the TI Processor, what will you do?

    Simple, port it to the new platform, its flexible! The project is not locked to the BBB, the system has already run on an iMX6, and other platforms.  In fact, it is all just part of the larger HAL_Linux layer of APM

    What is the PXF?

        The PXF is a cape for the BBB that has 3 IMU's onboard, a baro, and other support devices for the overall system to work.  It has protection for the processor to ensure that connecting devices like the Servos, do not feed back into the BBB and cause problems.

        The PXF project is platform agnostic, BBB was chosen to start with, we use the PRU's to ensure we have perfect timing for decoding and encoding servo and RX signals, this function of course could be ported to any small micro.  in fact with the rise of UAV-CAN, the need for PWM in future systems will diminish greatly

        So if you are interested in getting in at the ground floor, and riding the next wave of Autopilots, the PXF + BBB is a great start.  If you want the system to be more reliable, Jump on board, help!

       Victors work has been incredible, and without him, the project would not be where it is today.

    If you are not running Vision, what is the purpose of 1GHz?

    Thats simple.... more power for better flight characteristics.  as a system, APM moved from the APM 2.X series autopilot PCB's to the Pixhawk series, a jump from a 16MHz processor, to a 180 MHz processor within the last 2 years.  thats over 10 times the speed!  AND WE USE IT!

    so the jump to a BBB is only 5x faster than a Pixhawk!  we will use it!  I have never met a software engineer who has complained about too much memory, too much speed!

    The future for PXF?

    The PXF project will continue to evolve, along with the BBB, we are looking at other platforms, however, for now, we intend on staying with the BBB as the primary flight controller, and using ROS to integrate with ROS enabled Vision systems, SLAM, etc.

    Why is the name PXF™ Name Trademarked?

    We want you to know if the hardware you are purchasing has the active support of the team behind it.  The hardware is 100% OSHW, feel free to copy, modify etc etc... just respect the terms.... 

    However, if you want to purchase PXF™ hardware, and you want to know that the hardware you are buying is genuine, (and has all the correct hardware revisions in it), the trademark helps you.

    ERLE ROBOTICS uses the latest drawings from me, and they feed back any suggestions to the project.

    Contribute to diydrones/PXF development by creating an account on GitHub.
  • But Odroid is not open hardware.

  • odroid has limited io interfaces.. as an example, the u3 has bit-banged i2c :( recently with rev 0.5 there is an additional spi interface, which could be interesting..

This reply was deleted.