@Kashish-Garg-0 It's possible your board has a QSM8250 (M0054-2) instead of a QRB5165 (M0054-1), see here for the diffs: https://docs.modalai.com/m0054-versions/#m0054-1-and-m0054-2

If that is the case then you will have to QDL using a different build, which you can download from downloads.modalai.com, look for M0054-2 Starling QDL Image 1.7.4 2024-02-09-1.7.4-M0054-2.zip under the "All" category

@jameskuesel 30FPS currently. The AR0144 modules will outperform OV9782 modules as they have a better sensor and better lens

@Niek-van-Oostenbrugge We recommend replacing the files in the offline installer with your own

@Moderator @Zachary-Lowell-0 there is a similar issue with trying to include the message type AIDetection.h in ROS2 mpatoros2.

@Moderator Hi, do you know where can I get a description of the tsdf file format that the voxl-mapper uses?

@Darshit-Desai Sure, the easiest way would be to create a custom systemd service file

@Denver-Bennett Hi Denver, no voxl-tflite-server is a TensorFlow Lite runtime environment for VOXL.

You will need to write your own code if you want to somehow mesh models together over a network. We aren't familiar with any software that handles that today

@Zachary-Lowell-0 Yes I was able to install docker of u22.04, but at the same time when I was running everything on natively in melodic on the drone I was able to use 3rd party libraries like PCL ROS and I was able to install it, this was because the ROS installation was left on the user. I just thought it would be same with ROS2. Although now I have set up ARM64 dockers in my desktop ubuntu so now I can make as many changes on my desktop and just push the new docker

@Stefan-Amundarain Thanks for your patience! There is a slight bug in the flashing script that doesn't handle fixing the partition layout correctly without forcing it via. that arg. It'll be fixed in the next release (SDK 1.3..0)

@Manu-Verma , that is great that you figured it out!

Can you please help us understand where exactly the loose connection was? Could you please take a picture and mark the location of the issue? That would be really appreciated!

Alex

@anghung Okay I would retry flashing using QDL. If you still have issue you can submit an RMA and we can fix it for you: https://www.modalai.com/pages/rma

vox-tag-detector requires camera calibration, path to which which is stored here in the configuration file : /etc/modalai/voxl-tag-detector.conf . If you have not performed calibration of your tracking camera, please do that: https://docs.modalai.com/calibrate-cameras/

After that, make sure the voxl-tag-detector.conf is set up to listen to the correct camera (input_pipe should match the name of the camera that you are looking at through voxl-portal, typically tracking for the tracking caemra). Also make sure the detector is enabled "enable": true in voxl-tag-detector.conf .

Finally, as Tom suggested, if the tag detector encounters some error, it will display it in the log. You could just run the tag detector in the foreground by stopping the background service first and then running the detector in the terminal:

systemctl stop voxl-tag-detector voxl-tag-detector

@Stefan-Amundarain As a member of the software team I have no knowledge of your orders. I gave you the missing info you needed and a path forward using your existing hardware. If you'd like to get a replacement you can reach out to contact@modalai.com and they'll get you sorted quickly.

@Will-Anderson Yep, same deal. If you reach out to contact@modalai.com we can get you sorted. I'm passing this on to the operations team as well

@Denver-Bennett

Hey Denver, happy to try and help you out with this! Where you should start is our documentation for voxl-tflite-server which is how we often run ML models on VOXL. What you'll see from those docs is that we have a bunch of pretrained models already onboard as well as a bit of support for custom models. voxl-tflite-server then writes AI predictions out to a pipe using libmodal-pipe. However libmodal-pipe currently only has a C/C++ API and so you'd have to make a simple piece of C code to consume data from the pipe and invoke your Python script accordingly.

I'm more than happy to help with this, but it would be good to do some reading first. I'd check out that documentation link above as well as the one on the Modal Pipe Architecture to learn how processes communicate with each other. From there I can start to answer more specific questions and help you build this out.

Hope this helps!

Thomas Patton

@es24ucb , the M500 uses an off-the-shelf S500 V2 frame, which has 480mm diagonal distance between motors, so each motor is 240mm from the center. I hope this helps! I am sorry we do not have a CAD model of M500 that we can share at this time.

@rogerli said in Unable to communicate with add-on board:

@Vinny I have another VOXL2 where only DS1 and DS2 lights up constantly, but D1 doesn't light up. Is this normal? (There was a slight deformation near the power interface due to a previous fall.)

@tom The SDK version of the VOXL2 that I'm using is 1.1.3.

I initially bought two VOXL2 boards, both version 1.1.2. One was deformed due to a crash, but the other still works well. Later, I purchased two more VOXL2 boards, which arrived with version 1.1.3. One of them was completely non-functional, and the other also had communication issues. Due to time constraints, we urgently ordered two more VOXL2 boards. Should we downgrade the SDK to 1.1.2 for safety? Currently, as we are in the pre-mass production stage and sending products for inspection, we cannot afford sudden failures of critical components.

@ggiacalone I found something. When I start echoing a topic in voxl2, that topic (echoing) shows up on my laptop. I think we need to figure out the ros2 configuration of voxl.