3D Robotics

APM 2.0 beta released!

3689393223?profile=originalI'm delighted to announce the release of the first public beta of ArduPilot Mega 2.0. This is a significant upgrade to APM 1.02, and now delivers on full 2-way in-air communications, including real-time mission planning and autopilot control.


[UPDATE:Arduino-free code uploading added to features]


New features/changes include:

  • --Full support for the MAVlink protocol, which is now APM's default communications method.
  • --Auto-detection of GPS modules. No more requirement to tell APM what GPS module you're using--it will figure it out and configure the module appropriately!
  • 3689393150?profile=original--2-way telemetry.  Dozens of commands can now be sent from the Ground Station to the UAV while it's in the air. (see example from HK GCS at right)
  • --2-way Mission Planning. You can script and change missions in real time while the UAV is flying!
  • --Supports APM boards with the ATMega2560 chip, which has twice as much memory as the current board. The DIY Drones store will begin selling these new boards when the final version of APM 2.0 is released.
  • --Mission Planner now uploads firmware and checks for new versions online. No need for Arduino (unless you want it)!
  • --Magnetometer fully supported.
  • --Current sensor fully supported
  • --Support for the latest HK Ground Control Station and QGroundControl ground stations.
  • --Mission Planning and configuration can now be done wirelessly. No need to use USB if you don't want to.
  • --No more fiddling with a configuration file in Arduino! The software comes ready to go, right out of the box
  • --New command-line tests allow you to check failsafe operation, Xbees, radio inputs and more.
  • --New versions of HK GCS and the Mission Planner operate in MAVLink mode, allowing for wireless operation and compatibility with other MAVLink devices, including those running the Robot Operating System (ROS)


Huge props to the whole development team: Doug Weibel (project leader), Michael Smith, Jason Short, James Goppert, Michael Oborne (mission planner), Paul Mather (GCS), Lorenz Meier (MAVLink integration), Randy Mackay, Ryan Beall, Darren Corley, Joe Holdsworth, and countless others who have contributed with flight testing and bug catching!


Remember that this is an initial beta release, so there may be bugs and undocumented bits. After the beta-testing period (usually about six weeks) we'll release APM 2.0 final version, and retire APM 1.02.


Loading the code automatically:


Just download the Mission Planner and select "Upload Firmware" from the Tools section of the menu. Make sure APM is plugged into your USB port and you've selected the right port from the Mission Planner menu. Click on "Update Check" and the program will search online for the latest version of the code and load it to APM.


If you want to do a simulation with Xplane, follow the instructions below. That will soon be a firmware option with the Mission Planner, too.


Loading manually (Optional)

[UPDATE: Ignore this section if you don't want to fiddle with the code. Arduino is no longer required to load the code!]


Download the zip file, unzip it to your desktop. You will have two folders inside your APM 2.0 folder, as shown:



Make sure you're using the latest Arduino (0022). Set your Arduino Sketchbook location to the APM 2.0 folder, as shown below. You must exit Arduino and restart it for this to take effect. Within Arduino, open up the ArduPilotMega.pde file within the ArduPilotMega folder, and that will open all the other files. Before you compile, don't forget to select the right board (Arduino Mega 1280 or Arduino Mega 2560)



(Note: in the final version of APM 2.0, you will not have to use the Arduino IDE at all: the Mission Planner will load the firmware for you and check for more recent versions. The Arduino IDE will just be available in case you want to modify or inspect the code. UPDATE: this is now live)



Additional instructions:

--If you're doing flight simulation with Xplane, instructions for doing it with APM 2.0 are here.

--Make sure you're using the latest Mission Planner and HK GCS. In the Mission Planner, make sure you select APM 2.x modes (MAVLink), shown at right.


Known issues:

1) Mission Planner won't write waypoints on Port 0 in HIL mode. Instead, write the waypoints via your Xbee (port 3); remember to change the Mission Planner baud rate to 57k when using the Xbee port.
2) Uploading mission commands is still not fully documented. We'll get to that this week
3) There may be some issues with datalogs not recording. We're looking into it.

4) The manual still describes the 1.02 software. We'll change that to 2.0 during the beta test period, so by the time we release the final of 2.0 the manual is fully updated.

5) This has been mostly tested with 900 Mhz Xbee modules and the XstreamBee adapter, which work great. However, there is an issue with the 2.4 Ghz Xbee modules, which have a slightly different hardware configuration that conflicts with that adapter when sending data upstream. A short term workaround is to solder a jumper from the adapter's CTS to V+ pin. The next version of these adapter boards will include a more graceful fix.


If you find any bugs, please file a report in the Issue Tracker. The dev team will not be responding to bug reports filed in blog comments

E-mail me when people leave their comments –

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

Join diydrones


  • Krill, my GCS fix is ready for testing. 1.2.76.



  • Ok, thanks. That's good, I'm looking forward to test it.

  • No problem. I'm begging Doug to fix it today.....
  • Yes, It's likely you're right. I am not a native English speaker, so I can use sometimes inappropriate words. Sorry for this.

  • I made no such promise. It's not on my end. I've done what I can... but the read routine continues to bomb out due to a 1 second delay between waypoint messages (that I have not yet found in the APM source) and a 1 second timeout in MAVlink that does not recover when I send another request to re-send a waypoint.


    Deceptive is not the right word either. "Work around" would be appropriate. I'm not sure it's possible to deceive an in-intimate object like an APM.

  • Chris, yes I know about it. It was a deceptive solution. HappyKillmore told me about the root problem and promised that it will be fixed inthe next versions.
    I told the same in my instructional video (https://www.youtube.com/watch?v=cUVjoVaXPFM) and explained how to make it work temporarily.
  • 3D Robotics
    Kirill: Actually, any change to the HK GCS serial tab values will get the data flowing. It's a known issue and Paul is working on it. I've updated the manual to include this tip.
  • Any news on the 2.4 xbee and telemetry problems? I found out that the IMU board can read from Port3 if i hardwire-loop it and write out on Port3. Same with data from Xbee. Hardware looping RX/TX on xbee board causes successfull loopback. So it seems that its a rs232 voltage level issue. I found also that RX is working if TX is removed so that data from GCS to Ardupilot works if TX from ardupilot is disconnected..

    Btw; the workaround with CTS to VCC does not work for me..

  • Problem with simulating in XPlane using HK GCS via Xbee solved.

    The solution is here:


  • Thanks for the reply Chriss.

    I have read doug's post, The weird thing is that  I have had it working succesfully and turned the Ardupilot off. The next morning, It was back to square one. So It Can work, I will try to figure it out why it is intermittent. If I dont come right, I will have to resort to 868 xbees as South africa also uses 900mhz for 3G cellular connections.


This reply was deleted.