Hi @dlee - this is purposeful. The voxl2 uses a ton of dependencies on a specific build of opencv - so by installing ros-melodic-cv-bridge (which you should NOT even be able to do due to dependency conflicts) you break a lot of what is dependent on that build of opencv, including vision-hub. The ros melodic which is on target is the minimal build, meaning opencv isnt installed due to those specific conflicts above.

You can always try installing it from source and pointing the third party instead of downloading it in sudo. My other recommendation would be to upgrade to ros2 foxy which is already on target.

https://github.com/ros-perception/vision_opencv/tree/melodic <-- here is the link to to melodic cv bridge. LMK if that helps!

@cfirth said in Update SDK docs on installing ROS2 on VOXL2:

they

Thanks for the comment @cfirth - this has been updated in the gitlab and is waiting to be merged for the document update!

Makes sense to me @smilon - keep in mind that the px4 instance on the voxl2 is not mainlined so a lot of those features involving the microDDS arent fully implemented yet. It is quite easy to be the subscriber in ros2 from the DDS by exposing that topic from the yaml - but definitely might require some programming on the other side of things in PX4 to ensure that the rest of the code base can ingest the dds uorb you are passing along, etc.

Also just some stuff I have come accross with PX4 instance of precision landing - it isnt all that good - youd have better luck implementing your own method of landing via offboard movement using something like mavros or mavsdk.

Let me know how it goes!
Zach

Ok @jmltt I fixed it - there was a giant conflict with versioning on px4_msgs and px4_ros_coms which was fun to deal with - anyways I have a branch waiting for approval to be merged into the most recent nightly and you can then pull the dev from the apt repository when it gets merged in - until then feel free to rebuild voxl-mpa-to-ros2 deb and install on your voxl2 from the branch zach/submodule update: https://gitlab.com/voxl-public/voxl-sdk/utilities/voxl-mpa-to-ros2/-/tree/zach/submodule_update

Cheers!
Zach

@Judoor-0 this is a very unlikely occurance for a barebones instance with no external code running on the drone. I would recommend reflashing SDK 1.3.5 and the sys image and starting fresh to make sure there is nothing corrupt going on behind the scenes. Post install if this is still the case please let us know.

@smilon said in Custom px4-firmware build on StarlingV2:

I ran into an issue while installing the newly built voxl-px4 which states it depends on libfc-sensor(>=1.0.5) and I only have (1.0.4) on my system. @Eric-Katzfey I saw you authored some commits on libfc-sensor-api, do you have any clue how I can update?

1 Reply Last reply about 3 hours ago Reply

@smilon I would recommend using ros2 wrapper around mavsdk instead of mavros, that is what we have done in the past as mavros and ros1 are both depricated and not supported anymore. Anyways, goodluck with the project!

@vicky-s-karthik mind posting the exact commands that you ran?

ROS melodic is fully deprecated - so my recommendation is to move over to ros2 foxy which is what we support on the voxl2 as of right now. I will try and recreate your issue in the meantime. Please try installing ros melodic base as the cv2 instance on there conflicts with opencv on voxl-sdk.

Spot on @Matt-Ooi - the voxl2 is natively meant to handle transformations for ardupilot and px4 - BUT you are using MAVROS which is a different tranformation all together - hence why you would have to do the transofmratio yourself before sending into the flight stack running ardupilot (aka when calling mavros offboard) or something

@robertshmurda this has been done by myself by piping the point cloud directly into the MPA.

You can grab the point cloud from whatever api/sdk exists for your lidar and as long as you can get it into the correct struct to pass through the MPA, you can do that and stream it directly to voxl-mapper.

If you have more questions - feel free to email me zachary.lowell@ascendengineer.com and we can hop on a call or something.

Hi @Isaac-Stephan-Remy that tends to happen when the conversion from the frame in MPA to the ros1 or 2 image is not allocated the proper amount of size. I will do some testing with a FLIR thermal camera over UVC and get back with a fix for a PR - you will need to most likely build from source for ros1.

Also I do recommend swapping over to ROS2 FOXY on the voxl2 as it is more supported and ROS1 is at EOL.

Zach

@all - you can just change the SYS_ID of each drone and this should solve the port confliction.

@tomverstappen I am unable to recreate the issue with a starling running 1.5.0 - the output from voxl-vision-hub shows the relative coordinate frame from the tag relative to the camera and was able to get the drone to offset correctly.

Are there any logs you can paste in here or video?

Hello All - I will be trying to recraete this issue in the coming days - will keep you all posted.

Detections is pertinent to voxl-tflite-server and onboard computer vision on yolo, NOT on april tags.

@taiwohazeez I am not fully convinced you are running the right command still

Running voxl-inspect-tag shoudl show the following headers:

ID | XYZ | RPY | SIZE_M | CAM

As shown in the code below:
printf("\n"); printf("id:%3d name: %s\n", d.id, d.name); printf("XYZ: %6.2f %6.2f %6.2f \n", (double)d.T_tag_wrt_cam[0], (double)d.T_tag_wrt_cam[1], (double)d.T_tag_wrt_cam[2]); printf("RPY: %6.1f %6.1f %6.1f \n", (double)roll, (double)pitch, (double)yaw); printf("size_m: %4.2f latency_ms: %6.2f\n", (double)d.size_m, latency_ms); printf("cam: %s type: %s\n", d.cam, pipe_tag_location_type_to_string(d.loc_type)); printf("\n");

https://gitlab.com/voxl-public/voxl-sdk/services/voxl-tag-detector/-/blob/dev/clients/voxl-inspect-tags.c?ref_type=heads

Just run voxl-inspect-tags and that will output the right data.
Zach

Class confidence and detection confidence are pertinent towards voxl tflite server not voxl-tag-detector - are you sure this is the right output?

Hi @taiwohazeez - can you paste the output? Usually you get NANS or unknown when the locations.txt file isnt filled out meaning the tag detector service is referencing an empty file. Once you paste the output of voxl-inspect-detections I will be able to give you better guidance.

Thanks!
Zach

@Evan-Hertafeld said in Using MPA within voxl-docker-mavsdk-cpp:

libmodal_pipe

@Evan-Hertafeld @Jetson-Nano

You need to share all those files via a volume mount alongside with the /run/mpa directory to allow docker to probe the data and grab from an MPA pipe.

so in your docker run something like -v /usr/include/modal_start_stop.h:/usr/include/modal_start_stop.h needs to be done - the other option is you just download the actual MPA debian inside your docker image and call it a day - then you will have all the necessary headers, libs, binaries, and you just need to share /run/mpa as a volume to access the data.

Zach

@Jetson-Nano you could just use the CPP mavsdk instance to grab the TOF data from a MPA pipe and then leverage that to do your controls via mavsdk (I am assuming just offboard movement). You can connect over port 14551 just ensure you have localhost comms set to true in voxl-vision-hub.

@nikolahm is your drone PID tuned? How are you flying in position mode? GPS based flight, VIO based flight, etc. What is the environment you are testing in, can you paste a log in here please?