ROS driver for Parrot AR-Drone 1.0 and 2.0 quadrocopters
Home Page: http://wiki.ros.org/ardrone_autonomy
License: BSD 3-Clause "New" or "Revised" License
ROS Driver for Parrot AR-Drone 1.0 & 2.0 Quadrocopters
It isn't exactly related to the code, but here's an interesting question I hope one of you other researchers have answered before. When you send a command to the drone, for example cmd_vel.linear.x =1, are you controlling via velocity or acceleration?
The paper concerning this platform The Navigation and Control technology inside the AR. Drone micro UAV seems to go against the SDK.
The SDK 2.0 states that the input is "a percentage of maximum inclination as configured", while the paper says "velocity setpoints proportional to orientation of the handheld device"
What have you guys found in your work thus far? My lab is just beginning with the platform.
I recently installed the Ros Groovy and i'm having difficulties with the installation of the driver package. As mentioned, since JAn 9. groovy is supported, what does that mean?
I cant find the package.xml and nor the CMakeList.txt is in the new format, so I'm not able to compile with catkin_make the package.
The error is:
CMake Error at /opt/ros/groovy/share/catkin/cmake/catkinConfig.cmake:71 (find_package):
Could not find a configuration file for package camera_info_manager.
Set camera_info_manager_DIR to the directory containing a CMake
configuration file for camera_info_manager. The file will have one of the
following names:
camera_info_managerConfig.cmake
camera_info_manager-config.cmake
Call Stack (most recent call first):
ardrone_autonomy2/CMakeLists.txt:13 (find_package)
My package.xml downladable from: https://dl.dropbox.com/u/921645/package.xml
and the CMakeList.txt:
cmake_minimum_required(VERSION 2.4.6)
project(ardrone_autonomy)
include($ENV{ROS_ROOT}/core/rosbuild/rosbuild.cmake)
find_package(catkin REQUIRED COMPONENTS roscpp rospy std_msgs message_generation image_transport sensor_msgs std_srvs tf camera_info_manager)
catkin_package()
include_directories(include ${catkin_INCLUDE_DIRS} )
#set the default path for built executables to the "bin" directory
set(EXECUTABLE_OUTPUT_PATH ${PROJECT_SOURCE_DIR}/bin)
#set the default path for built libraries to the "lib" directory
set(LIBRARY_OUTPUT_PATH ${PROJECT_SOURCE_DIR}/lib)
#add_message_files()
#generate_messages()
find_package(Boost REQUIRED )
set(SDK ARDroneLib/)
link_directories(${PROJECT_SOURCE_DIR}/lib/)
include_directories(${SDK} ${SDK}/FFMPEG/Includes ${SDK}/Soft/Common ${SDK}/Soft/Lib ${SDK}/VP_SDK ${SDK}/VP_SDK/VP_Os/linux )
rosbuild_add_executable(ardrone_driver src/ardrone_driver.cpp src/video.cpp src/ardrone_sdk.cpp src/teleop_twist.cpp)
target_link_libraries(ardrone_driver pc_ardrone avcodec avutil swscale vlib sdk SDL )
How can I install it? Where can i find a version which is for the new catkin maker?
Hi all,
The messages of the Image and Camera info are not synchronized and that can cause problems to people working with that images...
My suggestion is this little modification in ardrone_driver.cpp (line 452):
sensor_msgs::Image image_msg;
sensor_msgs::Image::_data_type::iterator _it;
// Image and info msg need the same time stamp.
ros::Time camera_time = ros::Time::now();
image_msg.header.stamp = camera_time;
cinfo_msg_hori.header.stamp = camera_time;
cinfo_msg_vert.header.stamp = camera_time;
if (cam_state == ZAP_CHANNEL_HORI)
Thank you,
Josep Bosch
Ocean System Lab
Heriot-Watt University
Attempting to use the ROS package image_proc on the '/ardrone/front/image_raw' or '/ardrone/bottom/image_raw' steams fails to produce any corrected streams when used with the AR.Drone 2.0
The image_proc publishes the corrected streams, but gives an error indicating mismatched timestamps between images and the corresponding camera info when the stream is viewed.
The problem appears to occur due to lines 452-454 in ardrone_driver.cpp:
image_msg.header.stamp = ros::Time::now();
cinfo_msg_hori.header.stamp = ros::Time::now();
cinfo_msg_vert.header.stamp = ros::Time::now();
This doesn't guarantee the timestamps are matched as ros::Time::now() can change between calls.
The code was changed to:
image_msg.header.stamp = ros::Time::now();
cinfo_msg_hori.header.stamp = image_msg.header.stamp;
cinfo_msg_vert.header.stamp = image_msg.header.stamp;
This ensures timestamps are matched. The change appears to resolve the issue and image_proc is is able to correctly produce streams.
Whether or not this is also an issue for the original AR.Drone has not been tested.
Hello,
I am working with both the ardrone 1.0 and 2.0. When I run the ardrone_driver while connected to the drone 2.0 it runs fine. However, when I run the driver while connected to the drone 1.0 I get a message that it connected successfully with a 0% battery and then looses the nave data. The LED lights then turn red. This is all despite the fact that my battery was fully charged.
Below is what I see in my terminal:
Timeout when reading navdatas – resending a navdata request on port 5554
Timeout when reading navdatas – resending a navdata request on port 5554
Timeout when reading navdatas – resending a navdata request on port 5554
[INFO] [1353539677.060302095]: Successfully connected to 'My ARDrone' (AR-Drone 1.0 – Firmware: 1.7.11) – Battery(%): 0
Timeout when reading navdatas – resending a navdata request on port 5554
Hi,
I was wondering how the cmd_vel commands are relayed to the AR.Drone. When a linear-x is specified, is there an internal controller in the node that translates the desired linear-x to a pitch command to send to the drone? Or is setting linear-x the same as giving the drone a desired pitch (ie. is the command ardrone_at_set_progress_cmd(0, 0, 0.5, 0, 0) sent to the drone, or something of that nature?)? If there is some internal controller present, is there a way to directly send desired pitch and roll angles to the drone? Thank you!
Sarah
While flying an ardrone parrot 2.0, we are seeing a random drop in altitude. Watching the values from navdata_pwm, we noticed that the altitude_integral is resetting from 60.0 to 0.0 when the drop occurs. The integral term then rebuilds and the quad rises back to the desired altitude.
Has anyone else seen issues with this value resetting and if so how it can be fixed?
Hey!
I'm using an AR-Drone 2.0 with Firmware: 2.2.6. I try to switch between the bottom and front cam with $ rosservice call /ardrone/togglecam. But the result is that the /ardrone/front video feed is passed to the /ardrone/bottom so when i toggle the front video cam feed in on both topics and not different camera feeds on different topics.
I tired to launch the driver with the ardroen.launch, my.launch and with rosrun. Same result. The heli is on the ground, it works only after take off?
Am I missing something/ doing wrong, or this is a bug?
Hi all,
i am ashlih, i am newbie with ardrone and still learning.
i was wondering, is there a way to takeoff but with more slower speed,
so i can adjust the height of the first takeoff?
thank you very much for your answer..
Hi, I am trying to install your package in order to use the tum_ardrone package.
Unfortunately I can not build the AR Drone SDK: After I issued the build command following error is displayed:
Libs already extracted
Building target static
Architecture x86_64 is already built
Creating universal static lib file from architectures x86_64
Build done.
Checking required Ubuntu packages ...
ok.
Building ARDroneTool/Lib
ar rcs libvlib.a video_codec.o video_controller.o video_mem32.o video_dct.o video_huffman.o video_macroblock.o video_packetizer.o video_picture.o video_quantizer.o P263/p263_codec.o P263/p263_huffman.o P263/p263_mb_layer.o P263/p263_gob_layer.o P263/p263_picture_layer.o P264/p264_zigzag.o P264/p264_transform.o P264/p264_Qp.o P264/p264_intra_pred.o P264/p264_inter_mc.o P264/p264_merge.o P264/p264_codec.o P264/p264.o P264/p264_gob_layer.o P264/p264_mb_layer.o P264/p264_picture_layer.o P264/video_p264.o Stages/vlib_stage_decode.o Stages/vlib_stage_encode.o UVLC/uvlc_codec.o UVLC/uvlc.o UVLC/uvlc_gob_layer.o UVLC/uvlc_mb_layer.o UVLC/uvlc_picture_layer.o Platform/x86/video_utils.o Platform/x86/UVLC/uvlc_codec.o
make[5]: /home/hendrik/bin/ar: Kommando nicht gefunden
make[5]: *** [../../Soft/Build/targets_versions/vlib_PROD_MODE_Linux_3.5.0-21-generic_GNU_Linux_homehendrikbingcc_/libvlib.a] Fehler 127
make[4]: *** [all] Fehler 2
make[3]: *** [build_vlib] Fehler 2
make[2]: *** [all] Fehler 2
make[1]: *** [build_libs] Fehler 2
make: *** [all] Fehler 2
Could you please give some advice how to fix this?
I am not exactly a linux expert, so please forgive me if this error is trivial.
I am using ROS Groovy on an Ubuntu 12.10 machine.
I noticed that doing a flat trim before every take-off sometimes drastically improves the accuracy of the sonar data(as it is now calibrated to zero) and leads to more stable take-offs...Unfortunately I don't have the time now to figure out how pull requests work, so please consider adding this code to enable flat trims!
in ardrone_driver.cpp add
flatTrim_sub = node_handle.subscribe("ardrone/flatTrim",1,&flatTrimCallback);in ardrone_driver.h add
ros::Subscriber flatTrim_sub;
in `teleop_twist.cpp` add
void flatTrimCallback(const std_msgs::Empty &msg)
{
vp_os_mutex_lock(&twist_lock);
ROS_INFO("Setting flat trim now!");
ardrone_at_set_flat_trim();
ROS_INFO("Flat trim Set!");
vp_os_mutex_unlock(&twist_lock);
}user then publishes empty message to 'ardrone/flatTrim' topic to do a flat trim!
For some reason I am unable to fly either of my drones with the indoor or outdoor frame on the robot. The only way I can get it off the ground reliably is to fly the drone with out any of the covers over the battery.
Looking at the new motor pwm commands confirms that with the covers on the drone the motors are running about maxed out ~200 to 225. Without the frame, they are just below 200 and allow for full flight.
Has anyone else seen this problem or found a fix? Can anyone fly with the indoor frame on the robot?
Thanks-
parcon
I'm able to run the ardrone_drive properly and also control its movements using the cmd_vel topic.
When I do either rosrun image_view image_view image:=/ardrone/image_raw or rosrun image_view image_view image:=/ardrone/front/image_raw I can get a 640X360 pixel image.
But once I do rosrun image_view image_view image:=/ardrone/bottom/image_raw, I only get a blank window with no data..Right clicking gives a prompt saying "no data"!
Please help! I've made sure my firewall is disable while trying this, and have tested the bottom camera to be okay using both the Free Flight app and the Linux_sdk_demo...
Hi. Am using ardrone_autonomy package and trying to run the executable ardrone_driver
however, it is not working even after connecting to the drone. am getting the following error.
[ERROR] [1371144822.249433366]: [registerPublisher] Failed to contact master at [localhost:11311]. Retrying...
Can you please tell me wat is rong? Is it connection issue? I wanted to try this package, but am stuck here. please help me.
hi mani-monaj ,
first off all i will say this. great work that you have done.
it is for me the greatest, to look in the morning, if you have done something new.
the imu topic and also the camera calibration thing. all things that i want to do, but never had the time
or the experience to do this. fine to see how you solve all this things.
thx for all.
another question,
have you build launch files, where you use maybe things like robot_state_publisher or robot-pose-ekf
where i am able to see the drones doing in rviz. maybe like it was in the hector_quadrotor: packages.
or localize the drone via the tag detection?
this will be great if you publish something like that.
i build something like that with the starmac package, for the old drone.
you can see here.
[email protected]:mmyself/jh_ardrone.git
but this was yery rough.
all i really can recommend is the drone_teleop_3d i build,
great experience to fly the drone with a 3d mouse, because of the intuitive flying over the z-axis,
greats jan
When I try to access the video feed from the front camera I'm not getting a consistent video stream. It seems to publish (viewing using image_view or rviz) when there is not much detail on the screen (i.e. viewing my hand from up close or a constant color page) but it freezes (doesn't show anything) with complex images such as an outside scene or even a bookshelf with many colored books. I've tried to lower the bitrate by setting parameters 'bitrate_ctrl_mode' to 'VBC_MANUAL' and 'bitrate' to '1000' but that doesn't seem to change anything.
The bottom camera works smoothly with any complexity image.
It doesn't seem to be a problem with the camera as I get a reasonable picture using my iPhone and the FreeFlight app.
I tried to calibrate but the camera won't publish a picture as complicated as the checkerboard, I can get maybe 4 squares before it freezes again.
Running fuerte on Ubuntu 12.04 VM (hosted on windows 7)
Ardrone 1
Pulled the ardrone_autonomy git down on Friday (I believe)...not sure how to check the version
I doubt this is a bug in your code, but I'm wondering if you've seen this kind of behavior.
thanks
I am curently working to get multiple ardrones flying at the same time from a single desktop. I've built a simple launch file to create an instance of ardrone_autonomy in a namespace called "quad1". I also pass in ip arguments and a statement to handle which controller each drone is to be commanded from. My question is this: Even with compiling a second copy of ardrone_driver and attempting to run it along side the first instance of ardrone_autonomy, the second ardrone_driver fails. Does anyone know a way to resolve this error?
I see this error message:
[quad1/ARdrone-1] restarting process
process[quad1/ARdrone-1]: started with pid [17386]
Using custom ip address 192.168.1.10
Wait authentification
Wait authentification
Wait authentification
===================+> 192.168.1.10
[ INFO] [1360280789.946051099]: Root Frame is: 0
[ INFO] [1360280789.947098353]: ~cov/imu_la set to [0.1, 0, 0, 0, 0.1, 0, 0, 0, 0.1]
[ INFO] [1360280789.948039719]: ~cov/imu_av set to [1, 0, 0, 0, 1, 0, 0, 0, 1]
[ INFO] [1360280789.949083298]: ~cov/imu_or set to [1, 0, 0, 0, 1, 0, 0, 0, 100000]
[ WARN] [1360280789.949989201]: Automatic IMU Caliberation is active.
[ INFO] [1360280789.950096522]: using default calibration URL
[ INFO] [1360280789.950166558]: camera calibration URL: file:///home/parcon/.ros/camera_info/ardrone_front.yaml
[ INFO] [1360280789.953101752]: using default calibration URL
[ INFO] [1360280789.953193572]: camera calibration URL: file:///home/parcon/.ros/camera_info/ardrone_bottom.yaml
Deleting Profile -cea48d22
Deleting Application -e182b69b
Input device Teleop added
Starting thread video_stage
Starting thread update_ros
video stage thread initialisation
Video multisocket : init 2 sockets
Video multisocket : connecting socket 0 on port 5555 UDP
Thread update_ros started
video stage thread ended
Academy download stage paused
[ INFO] [1360280795.015007096]: SEND: CAT_COMMON/altitude_max = 2500.000000 (DEFAULT = 3000.000000)
[ INFO] [1360280795.015084296]: SEND: CAT_COMMON/enemy_colors = 2.000000 (DEFAULT = 1.000000)
[ INFO] [1360280795.015169021]: SEND: CAT_APPLI/navdata_options = 268435456.000000 (DEFAULT = 65537.000000)
[ INFO] [1360280795.015248827]: SEND: CAT_USER/euler_angle_max = 0.087000 (DEFAULT = 0.209440)
[ INFO] [1360280795.015302954]: SEND: CAT_USER/control_yaw = 1.750000 (DEFAULT = 1.745329)
[ INFO] [1360280795.015363624]: SEND: CAT_APPLI/bitrate = 4000.000000 (DEFAULT = 1000.000000)
[ INFO] [1360280795.015419406]: SEND: CAT_SESSION/max_bitrate = 4000.000000 (DEFAULT = 1000.000000)
[ INFO] [1360280795.015475310]: SEND: CAT_SESSION/detections_select_h = 32.000000 (DEFAULT = 0.000000)
[quad1/ARdrone-1] process has died [pid 17386, exit code -11].
log files: /home/parcon/.ros/log/7240ab1e-717e-11e2-a955-1c6f65c6b0bc/quad1-ARdrone-1*.log
respawning...
[quad1/ARdrone-1] restarting process
process[quad1/ARdrone-1]: started with pid [17607]
Using custom ip address 192.168.1.10
So according to this paper, (http://cas.ensmp.fr/~petit/papers/ifac11/PJB.pdf) the navdata velocity estimates are based on fused estimates from the optical flow and the aerodynamic model of the drone. When an estimate is unavailable from the bottom camera, simply the aerodynamic model is used.
This means the biases are recalculated automatically as and when the estimate switches from a fused one to the aerodynamic one. So, does it make sense for you to still do your recalibration in the beginning and stick with that value through a given run of the drone? Wouldn't that already be done much better on the embedded system?
The hover state takes into account features on the image plane to keep the ardrone stationary, but I would like to be able to hover without using the cv methods. So far I believe that using very small velocities in the cmd_vel/linear message is the best way to accomplish this.
According to the SDK this is a simple bit value that is sent with the command velocity. Is their a way to set this bit manually? Maybe the cmd_vel/angular/x or cmd_vel/angular/y could be high-jacked for this function, since they are not in use.
Hello,
...After getting a successful Rosmake of the package, when I tried to run the driver it emits a segfault during the initialization process. I'm not sure what exactly is going and would really appreciate some help.
Thanks,
Sen
Ubuntu 12.10
Ros Groovy
heres the output...
user$~ rosrun ardrone_autonomy ardrone_driver
Wait authentification
Wait authentification
Wait authentification
===================+> 192.168.1.1
[ INFO] [1368262008.957043787]: Root Frame is: 0
[ INFO] [1368262008.958842421]: No values found for ~cov/imu_la in parameters, set all to zero.
[ INFO] [1368262008.960211071]: No values found for ~cov/imu_av in parameters, set all to zero.
[ INFO] [1368262008.961594793]: No values found for ~cov/imu_or in parameters, set all to zero.
Deleting Profile -cea48d22
Deleting Application -e182b69b
Input device Teleop added
Starting thread video_stage
Starting thread video_recorder
video stage thread initialisation
Starting thread update_ros
Video multisocket : init 2 sockets
Video multisocket : connecting socket 0 on port 5555 UDP
Video recorder thread initialisation
Video multisocket : connecting socket 1 on port 5555 TCP
Set IP_TOS ok
Starting thread navdata_update
Thread update_ros started
Starting thread ardrone_control
Thread navdata_update in progress...
Start thread thread_academy_download
Academy download stage resumed
PA : MEMORY SPACE ALLOWED : 40 MB
Start thread thread_academy
Start thread thread_academy_upload
Segmentation fault (core dumped)
I'm using the Ar.Drone 2.0. with Firmware: 2.2.6.
The front camera can be used with resolution : 360p (640x360) or 720p (1280*720). The default setup is the first one, 360p. I would like to use the 720p.
How can I change it? Is there a parameter in the launch file?
Is there any update in camera togle issue?
Hello everyone,
I would like to acquire images from both AR Drone cameras as fast as possible. To do so I have programmed a test code using ROS (using the ardrone_autonomy driver https://github.com/AutonomyLab/ardrone_autonomy , which in turn uses the AR Drone SDK). From what I have seen, a fair ammount of images can be received (around 10-20) when switching between cameras. The problem is that part of the images received from the frontal camera are in reality from the bottom camera; and the same happens the other way around. For instance when I visualize the video streams "ardrone/front/image_rect_color" and "ardrone/bottom/image_rect_color" on two separate "image_view" nodes; I will get images from both cameras on both "image_view" windows. Has anybody encountered the same problem before?
I am using the latest version of the SDK and the ardrone_autonomy software on Ubuntu 12.04 and an AR Drone 2.0.
Please note that there other users interested on this issue:
https://projects.ardrone.org/boards/1/topics/show/1609
https://projects.ardrone.org/issues/show/172
My question is if other people are facing the same problem, and whether this bug is a problem on the SDK or on the ardrone_autonomy node.
Thanks for your help,
jespestana
I'm using roscpp, and can't seem to get the drone to take off unless I enable latching on the takeoff publisher:
ros::Publisher launch = n.advertise<std_msgs::Empty>("ardrone/takeoff", 5, true);
ros::Duration(1.0).sleep();
std_msgs::Empty empty;
launch.publish(empty);Even then (and with or without latching on the land publisher) I can't get the drone to land from roscpp.
However, command line works great for takeoff landing and reset:
$ rostopic pub -1 /ardrone/takeoff std_msgs/EmptyWhy?
It appears that changes in control parameters (and other CAT_USER) settings do not affect the drone.
When running ardrone_driver, the following is printed:
Wait authentification ===================+> 192.168.1.1 [ INFO] [1352234370.655983884]: Root Frame is: 0 [ INFO] [1352234370.656905814]: No values found for `~cov/imu_la` in parameters, set all to zero. [ INFO] [1352234370.657902223]: No values found for `~cov/imu_av` in parameters, set all to zero. [ INFO] [1352234370.658896658]: No values found for `~cov/imu_or` in parameters, set all to zero. [ INFO] [1352234370.663244082]: SET ~bitrate_ctrl_mode : 0.00 [ INFO] [1352234370.664260742]: SET ~max_bitrate : 2000.00 [ INFO] [1352234370.665207506]: SET ~outdoor : 1.00 [ INFO] [1352234370.666118191]: SET ~flight_without_shell : 1.00 [ INFO] [1352234370.667830458]: SET ~altitude_max : 5000.00 [ INFO] [1352234370.668715842]: SET ~altitude_min : 100.00 [ INFO] [1352234370.670661813]: SET ~enemy_colors : 3.00 [ INFO] [1352234370.671968513]: SET ~enemy_without_shell : 0.00 [ INFO] [1352234370.673430208]: SET ~control_vz_max : 2000.00 [ INFO] [1352234370.674487957]: SET ~control_yaw : 350.00 [ INFO] [1352234370.675655010]: SET ~euler_angle_max : 0.50 [ INFO] [1352234370.676419481]: SET ~bitrate : 2000.00 [ INFO] [1352234370.677528147]: SET ~navdata_demo : 0.00 [ INFO] [1352234370.678744205]: SET ~detect_type : 10.00 [ INFO] [1352234370.679771770]: SET ~detections_select_h : 32.00 [ INFO] [1352234370.680836172]: SET ~detections_select_v_hsync : 4.00 Input device Teleop added Starting thread video_stage Starting thread video_recorder Starting thread update_ros video stage thread initialisation Video multisocket : init 2 sockets Video multisocket : connecting socket 0 on port 5555 UDP Video multisocket : connecting socket 1 on port 5555 TCP Thread `update_ros` started Video recorder thread initialisation Set IP_TOS ok Starting thread navdata_update Starting thread ardrone_control Thread navdata_update in progress... PA : MEMORY SPACE ALLOWED : 40 MB Start thread thread_academy Start thread thread_academy_upload Start thread thread_academy_download Academy download stage resumed Academy download stage paused Timeout when reading navdatas - resending a navdata request on port 5554 Academy download stage resumed Academy download stage paused Sending default CAT_SESSION settings [ INFO] [1352234375.681546174]: Successfully connected to 'My ARDrone' (AR-Drone 2.0 - Firmware: 2.1.20) - Battery(%): 57
Investigating the drone's configuration files (see next code snippet) shows that the CAT_SESSION values (for example detect_type, detections_select_h, detections_select_v_hsync) are correctly set (as implied by the driver output), but custom CAT_USER settings are not set. The type of each parameter can be read in config_keys.h.
CAT_SESSION parameters correctly set
# pwd /data/custom.configs/sessions # cat config.7bc641ad.ini [control] flying_mode = 0 hovering_range = 1000 [video] codec_fps = 30 video_codec = 129 max_bitrate = 2000 video_channel = 0 [detect] groundstripe_colors = 16 detect_type = 10 detections_select_h = 32 detections_select_v_hsync = 4 detections_select_v = 0 [userbox] [gps] latitude = 5.0000000000000000e+02 longitude = 5.0000000000000000e+02 altitude = 0.0000000000000000e+00 [custom] application_id = e182b69b profile_id = 00000000 session_id = 7bc641ad session_desc = Session 7bc641ad
But CAT_USER not?
# cd ../profiles/ # ls
There are no user profiles created by the application.
Further debugging by setting #define _GENERAL_NAVDATA_DEBUG (1) in ardrone_general_navdata.c produces the following output on driver startup:
Wait authentification ===================+> 192.168.1.1 [ INFO] [1352234370.655983884]: Root Frame is: 0 [ INFO] [1352234370.656905814]: No values found for `~cov/imu_la` in parameters, set all to zero. [ INFO] [1352234370.657902223]: No values found for `~cov/imu_av` in parameters, set all to zero. [ INFO] [1352234370.658896658]: No values found for `~cov/imu_or` in parameters, set all to zero. [ INFO] [1352234370.663244082]: SET ~bitrate_ctrl_mode : 0.00 [ INFO] [1352234370.664260742]: SET ~max_bitrate : 2000.00 [ INFO] [1352234370.665207506]: SET ~outdoor : 1.00 [ INFO] [1352234370.666118191]: SET ~flight_without_shell : 1.00 [ INFO] [1352234370.667830458]: SET ~altitude_max : 5000.00 [ INFO] [1352234370.668715842]: SET ~altitude_min : 100.00 [ INFO] [1352234370.670661813]: SET ~enemy_colors : 3.00 [ INFO] [1352234370.671968513]: SET ~enemy_without_shell : 0.00 [ INFO] [1352234370.673430208]: SET ~control_vz_max : 2000.00 [ INFO] [1352234370.674487957]: SET ~control_yaw : 350.00 [ INFO] [1352234370.675655010]: SET ~euler_angle_max : 0.50 [ INFO] [1352234370.676419481]: SET ~bitrate : 2000.00 [ INFO] [1352234370.677528147]: SET ~navdata_demo : 0.00 [ INFO] [1352234370.678744205]: SET ~detect_type : 10.00 [ INFO] [1352234370.679771770]: SET ~detections_select_h : 32.00 [ INFO] [1352234370.680836172]: SET ~detections_select_v_hsync : 4.00 Input device Teleop added Starting thread video_stage Starting thread video_recorder Starting thread update_ros video stage thread initialisation Video multisocket : init 2 sockets Video multisocket : connecting socket 0 on port 5555 UDP Video multisocket : connecting socket 1 on port 5555 TCP Thread `update_ros` started Video recorder thread initialisation Set IP_TOS ok Starting thread navdata_update Starting thread ardrone_control Thread navdata_update in progress... PA : MEMORY SPACE ALLOWED : 40 MB Start thread thread_academy Start thread thread_academy_upload Start thread thread_academy_download Academy download stage resumed Academy download stage paused Timeout when reading navdatas - resending a navdata request on port 5554 [286] General Navdata : Need to check multiconfig ... request config file Academy download stage resumed Academy download stage paused [69] General Navdata : Config callback called with result 1 [72] General Navdata : State : 6 [76] General Navdata : Got config file [102] General Navdata : End of config callback call [166] General Navdata : Checking drone version ... [170] General Navdata : Drone supports multiconfig [69] General Navdata : Config callback called with result 1 [72] General Navdata : State : 7 [80] General Navdata : Got ids list [102] General Navdata : End of config callback call [184] General Navdata : Got AR.Drone ID list. Switch->1,1,1. ND->1 [198] General Navdata : Checking application efc8fd84 (desc : com.parrot.freeflight) [198] General Navdata : Checking application b6bc397b (desc : com.digitalsirup.dronecontrolhd) [198] General Navdata : Checking application e182b69b (desc : ardrone_driver) [201] General Navdata : Found our application ... should not init [234] General Navdata : We're requesting default user (00000000), do nothing. Sending default CAT_SESSION settings [69] General Navdata : Config callback called with result 1 [72] General Navdata : State : 8 [84] General Navdata : Got current ids [102] General Navdata : End of config callback call [292] General Navdata : Send application navdata demo/options [69] General Navdata : Config callback called with result 1 [72] General Navdata : State : 9 [88] General Navdata : Got navdatas [102] General Navdata : End of config callback call [301] General Navdata : Refreshing from MULTICONFIG FSM [ INFO] [1352234375.681546174]: Successfully connected to 'My ARDrone' (AR-Drone 2.0 - Firmware: 2.1.20) - Battery(%): 57
The driver seems to use the default user profile, which is not updated according to the set ros parameters.
Changing the euler_angle_max parameter to 30º instead of the default 12º and running the drone in a vicon system shows no change in drone behavior with the increased parameter value (this is also visually apparent).
Perhaps I am doing something incorrectly, in which case please let me know! If however this is an issue, I am happy to help debug – let me know what I can do!
Mike
This issue is related to #39 & #41
First I restate one of my previous comments with modifications:
I did more tests on "realtime" and "legacy" publishing of "Navdata" and here are my thoughts:
"Full Speed Navdata" is not a good idea. The current implementation bypasses main ROS loop running in a different thread while instantly publishes the data which may generate undefined behaviour. I say this because I could not find any information about ROS that states it is safe to bypass the same ROS loop from another thread.When navdata_demo is turned on and the Drone is publishing at 200Hz, the main ROS loop can not catch up even with 500Hz loop rate. I think this is because of expensive memory copies done in video and navdata callback functions, many locks/unlocks between two threads and other custom code/ROS overheads during the ROS loop.
In order to examine it more, I put a ROS_INFO inside this if block to print the difference between copy_current_frame_id and last_frame_id. If ROS loop is in sync with drone update loop, this number should be always 1, otherwise it shows the number of navdata misses between consequent ROS updates which consequently leads to frame rates less than 200Hz. The result was as I expected. There are lots of misses when the looprate is set to 200Hz. Increasing that number will lead to better navdata frame rate (and less misses). In my case setting that number to 1000Hz would lead to 200Hz navdata update rate.
In summary, I suggest to remove the "Full Speed Navdata". Anyone who is interested in getting 200Hz navdata update, can increase the looprate variable appropriately.
The current driver thread model is that ROS update loop (ARDroneDriver::run()) is being run in another thread alongside SDK2's video_thread and navdata_update using SDK2 threading macros. The problem of the ROS thread being unable to catch up with fast (200Hz) navdata_update is that the ROS loop has a wait inside it. While ROS loop is [busy] waiting by pre-defined frequency (loop_rate.sleep()), it will eventually miss a couple of data frames that is being received in the other thread in navdata_custom_process.
The reason why the fullspeed_navdata works is that it pushes the new data into ROS publish buffer for a topic which will be pushed to topic's queue on ros::spinOnce() looprate times in a second. I was wondering if it is a good idea that we re-structure the driver's code so that publish_navdata and publish_video is being called directly from SDK2 thread exactly the same way that rosDriver->PublishNavdataTypes(shared_raw_navdata) is being called now.
My concern is whether publishing to ROS from another thread is safe or not. As far as I could understand, the ROS multi-threaded spinning model is meant for subscribers/publishers from different thread, which is slightly different from our case.
I would really appreciate any comments @mikehamer, @JakobEngel.
Hey,
Can anybody help me what exactly does this wind_comp_angle do? Can I stabilize the drone with this angles? If yes, how to use it ?
I also notice that while changing height only in Free flight drone also compensates for the wind, so that the drift is not to big. If using ROS and it's progressive commands, the drift is not compensated.
Any thoughts, or help?
thank's
Grega
Are the velocity measurements just the vision based estimates or are they a fused measurement incorporating accelerometer data?
The new commit makes the stability much better than the previous one, thanks a lot for that! However everytime I give the drone a yaw or throttle command, the drift is way too much to ignore. Roll and Pitch is fine, and hovering is very good now. Of course, I tested the drone in the same configuration with the official apps, and yaw and throttle are both alright then...
So, is there a reason why this could be happening?
The 2.0 drone sends timestamps with each videoframe (parrot_video_encapsulation_t field timestamp).
It would be extremely helpful for me to get these somehow attached to each frame, however that's not thet easy, as they get lost somewhere in the parrot video-processing pipeline (i.e. would have to add a driver_pre_stages or hack around in the SDK). Also, these timestamps are not in sync with the navdata timestamps, i dont know how you could synchronize them.
Hi, i am not sure why and if it is really in issue.
But after pulling the new updates i have no video more.
any ideas what can go wrong there.
is it possible the no video subscriber thing make the trouble for me?
i only check if the video is there with
rosrun image_view image_view image:=/ardrone/image_raw
with a reset to
git reset 58d20cd
it works again.
greets jan
There's some typos in the Navdata section in Readme.MD:
The lines:
rotx: Left/right tilt in degrees (rotation about the X axis)
roty: Forward/backward tilt in degrees (rotation about the Y axis)
rotz: Orientation in degrees (rotation about the Z axis)
Should be rotX, rotY and rotZ (capital letter for axis name)
I0m starting to use ROS and this library, and I made this simple program to lift the ardrone more than a certain threshold (threshold = umbral in my program). First I send and empty message and to /ardrone/takeoff , it take offs, and the when I start running my program, the ar drone looses altitude and it is not able to reach an altitude greater than 40 cm or so. I'll really appreciate tips or advices... I'm new at this stuff...
Some defines and variables are not used, i declared them thinking in using them afterwards
float altitud;
float rotx;
float roty;
float rotz;
float umbral = 400;
float vz=0;
float ax=0;
float ay=0;
bool flag = true;
void navDataCallback(const ardrone_autonomy::Navdata::ConstPtr& navigationData){
if (flag){
umbral_rotz = navigationData->rotZ;
umbral = navigationData->altd + umbral;
flag = false;
}
altitud = navigationData->altd;
ROS_INFO("%f \t %f \t error: %f", altitud, umbral, umbral-altitud);
}
int main(int argc, char ** argv){
ros::init(argc, argv, "ardrone");
ros::NodeHandle n;
ros::Subscriber navDataSub = n.subscribe("/ardrone/navdata", 100, &navDataCallback);
ros::Publisher pub = n.advertise <geometry_msgs::Twist> ("/cmd_vel",1000);
ros::Rate loop_rate(10);
while(ros::ok){
geometry_msgs::Twist msg;
if (altitud>umbral+200){
vz = -1.0;
}
else if (altitud<umbral){
vz = 1.0;
}
else{
vz = 0;
}
msg.linear.z = vz;
pub.publish(msg);
ros::spinOnce();
loop_rate.sleep();
}
return 0;
}
In the documentation, it says to call the service ardrone/flattrim. After running the driver successfully, the service does not exist, and after searching the directories I couldn't find it. Has it been implemented yet?
I'm trying to implement autonomous landing using the bottom camera, and this application heavily relies on a steady, constant camera frame rate. The current frame rate can vary from anywhere between 0.1-0.12 secs, and sometimes when it freezes, there is no image for up to 2 secs altogether. although I don't need the front camera now, the stream from it seems to be fairly constant, but the bottom camera gives a very choppy stream more often than not.
I've tested this with my freeflight app of course, and it seems fine then.
This feed is usually accompanied by messages like "Timeout when reading navdatas - resending a navdata request on port 5554", "Decoding failed for I frame", "Decoding Failed for P frame". What can i do to improve the link between the computer and the drone? I'm guessing that's the cause for this, and not a bug in the driver?
Please forgive me if this is the wrong place to ask this. I am working with the ardrone for my senior thesis and ardrone_autonomy looks absolutely awesome. Thanks for your hard work.
I was wondering if you could check out the issue I am having. When I try and run rosrun ardrone_autonomy ardrone_driver, I get the following output, ending in "The AR-Drone has a suspicious Firmware number."
https://gist.github.com/3825341
As far as stats, I am running Ubuntu 12.04 and have an ardrone 2. I would appreciate any help you could offer. Thanks!
Does the ardrone_driver expose the interface to turn on and off video recording to the USB drive? I noticed the video_on_usb parameter, which defaults to 1, but I connected a USB drive to my AR.Drone and recording did not happen. I also quickly reviewed the AR.Drone SDK documentation and source code, and I couldn't find a call in there which enabled recording to the USB. Any ideas?
Thanks
-s
The tread has seemed a little dead lately....so how about another question- I've been having trouble getting multiple drones working with the drivers. Has anyone found a good method for multiple drones? I'm expecting to use multiple wireless cards, and if necessary, some custom static routes.
I have problems when trying to import "from ardrone_autonomy.msg import NavData" with python.
Could you help me?
Hey @mani-monaj,
Just wondering what the timeline is for merging the dev-unstable changes (eg full-speed navdata and extra navdata packets) into the master branch?
Cheers,
Mike
Hi! First of all, thank you for taking the time to create such a helpful package, I''m finding it very useful for my senior thesis project. One point of clarification - the instructions say to call the togglecam service with an empty message. There's additionally a CamSelect.srv service messages as well. What is the difference between the two? Thank you!
I just tried ti build the ardrone_driver in ROS Groovy. In Groovy some things concerning the bullet engine have been changed[1]. That's why the driver won't build (btQuaternion not defined). Here's my proposed patch:
--- ardrone_driver.cpp 2012-12-21 15:17:32.706078979 +0100
+++ ardrone_driver.cpp.old 2012-12-21 15:27:34.022090067 +0100
@@ -2,7 +2,6 @@
#include "teleop_twist.h"
#include "video.h"
#include <signal.h>
-#include <tf/LinearMath/Quaternion.h>
////////////////////////////////////////////////////////////////////////////////
// class ARDroneDriver
@@ -699,7 +698,7 @@
imu_msg.linear_acceleration.z = msg.az * 9.8;
// IMU - Rotation Matrix
a very small issue. not really important.
you wrote in the readme that the default value is 1
enemy_without_shell - Default: 1
but in the code it is 0.0
ardrone_application_default_config.enemy_without_shell = (bool) rosDriver->getRosParam("~enemy_without_shell", (double) 0.0);
Hi folks! I've put the examples mentioned in my former push request into it's own git. The repo can be found here: https://github.com/parcon/arl_ardrone_examples . I'll have it added to the ROS index today also.
I installed new release of ROS (Groovy) and when I compile ardrone_autonomy rosmake returned an
error. I resolve by changing ardrone_driver.cpp like below:
// IMU - Rotation Matrix
btQuaternion q;
q.setEulerZYX(msg.rotZ * _DEG2RAD, msg.rotY * _DEG2RAD, msg.rotX * _DEG2RAD);
tf::quaternionTFToMsg(q, imu_msg.orientation);
---->
// IMU - Rotation Matrix
tf::Quaternion q;
q.setEulerZYX(msg.rotZ * _DEG2RAD, msg.rotY * _DEG2RAD, msg.rotX * _DEG2RAD);
tf::quaternionTFToMsg(q, imu_msg.orientation);
I'm fairly new to ROS so I may be wrong here, but shouldn't the commands for controlling the drone be published to cmd_vel topic rather than the ardrone/cmd_vel topic? The manual says the publish on the second topic, which basically does nothing...publishing on the first one works though!
Hi Guys,
Firstly I want to thank you for releasing this driver publicly, it's been a huge assistance to me.
I've been trying the last few days to enable shell-tag detection from the front camera of my AR.Drone 2.0, but unfortunately without success. The following parameters are set to the drone:
SET ~bitrate_ctrl_mode : 0.00
SET ~max_bitrate : 4000.00
SET ~outdoor : 1.00
SET ~flight_without_shell : 1.00
SET ~altitude_max : 5000.00
SET ~altitude_min : 100.00
SET ~enemy_colors : 3.00
SET ~enemy_without_shell : 0.00
SET ~control_vz_max : 2000.00
SET ~control_yaw : 350.00
SET ~euler_angle_max : 0.50
SET ~bitrate : 4000.00
SET ~navdata_demo : 0.00
SET ~detect_type : 10.00
SET ~detections_select_h : 32.00
SET ~detections_select_v_hsync : 128.00
Where:
enemy_colors = 3 = ARDRONE_DETECTION_COLOR_ORANGE_BLUE (my tag color)
detect_type = 10 = CAD_TYPE_MULTIPLE_DETECTION_MODE
detections_select_h = 32 = TAG_TYPE_SHELL_TAG_V2
detections_select_v_hsync = 128 = TAG_TYPE_BLACK_ROUNDEL
Detection of the black roundel from the bottom camera works successfully and reliably.
Do you have any suggestions about what I could be doing wrong?
Thanks and best regards,
Mike
i run in Ubuntu 10.04 64bit in this little issue by rosmake ardrone_autonomy
In file included from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/FFMPEG/Includes/libavutil/avutil.h:125,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/VP_SDK/VP_Api/vp_api_picture.h:16,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/VLIB/video_controller.h:5,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/VLIB/video_codec.h:4,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/Soft/Common/config_keys.h:29,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/Soft/Common/ardrone_api.h:739,
from /home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/Soft/Lib/ardrone_tool/ardrone_tool.h:4,
from /home/jan/ros/stacks/ardrone_autonomy/src/ardrone_sdk.h:25,
from /home/jan/ros/stacks/ardrone_autonomy/src/ardrone_driver.h:8,
from /home/jan/ros/stacks/ardrone_autonomy/src/ardrone_driver.cpp:1:
/home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/FFMPEG/Includes/libavutil/common.h: In function ‘int32_t av_clipl_int32_c(int64_t)’:
/home/jan/ros/stacks/ardrone_autonomy/ARDroneLib/FFMPEG/Includes/libavutil/common.h:179: error: ‘UINT64_C’ was not declared in this scope
solved it with this small code, found by google
wonder why. because the SDK2.0 compile in this system well, also the examples.
maybe a make flag thing in ros???
Hi,
I'm pretty new with ROS so please, bear with me.
When I entered
rosdep install ardrone_autonomy
into the terminal, I got
ERROR: the following packages/stacks could not have their rosdep keys resolved to system dependencies:
ardrone_autonomy: Cannot locate rosdep definition for [libsdl-dev]
I'm currently using Ubuntu 12.04 with ROS Fuerte. I checked to make sure libsdl-dev was installed on my OS, and it said it has libsdl1.2-dev. I also tried changing the manifest.xml file in ardrone_autonomy to say libsdl1.2-dev instead of libsdl-dev, but that didn't help.
Can I please get some help on this issue? Thanks!
EDIT:
I've already looked at
http://answers.ros.org/question/35426/problem-with-linking-sdl-when-compiling-ardrone_brown-node/
They had the same problem, but the version of ardrone_autonomy I'm using already has SDL at the end of the target_link_libraries(...).
I apologize if I am posting this issue in the wrong area. When I enter:
$ rosmake ardone_autonomy
I get the following message:
mkdir: cannot create directory 'build': Permission denied
I have tried the follow to fix this:
$ sudo -i
$./build_sdk
which executes, but when I go to rosmake ardone_autonomy I still get the same error. Is there something simple I overlooked in the tutorial? I am using the VirtualBox virtual image with Ubuntu 12.04.1 LTS and ROS Fuerte pre-installed. If this is the incorrect location for this issue, could anyone point me to the correct forum. Thank you
So I'm using the April Tag Detector for the autonomous landing operation, and that requires me to specify a url of the camera I want it to stream the images from. For example, the URL of my webcam is
v4l2:///dev/video0
Is there a way I can specify the url of the drone's bottom camera to the application?
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
Django
The Web framework for perfectionists with deadlines.
Laravel
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
Microsoft
Open source projects and samples from Microsoft.
Google
Google ❤️ Open Source for everyone.
Alibaba
Alibaba Open Source for everyone
D3
Data-Driven Documents codes.
Tencent
China tencent open source team.