Pixhawk 2 and Jetson TX1?


Is there anyone that have successfully used the Pixhawk 2 with a Jetson TX1 companion?


Hi Ricardo,

yes we are successfully using Pixhawk 2 with NVIDIA Jetson TX1 + our own designed carrier board for it. You can find more information about PIXEVIA CORE X1 in question here - http://www.pixevia.com/p-core?i=s50.

If you have any further questions regarding the Jetson TX1 use case via Pixhawk and drones, feel free to ask here or contact us directly.

1 Like

Hello @PauliusV and thanks for your reply. I read it just now because I had a break from the project.

Now we are back on it and we are going to purchase the HW. I will let you know if we need any kind of advice or collaboration from you. Actually, your HW and SW looks a good option for us. Let’s get in touch.

In the meantime, all the best.

This sounds very interesting. I’m currently using a Raspberry Pi 3 communicating with the Pixhawk using Dronekit and opencv for feature recognition. I want to move towards the use of neural networks and consider the Pi somewhat limited for this purpose, despite its great flexibility.

The Jetson is the natural choice, but I’m not sure what software combination to use. Dronekit or ROS? Caffe or Tensorflow?

All thoughts welcome.

how about disclosing some idea of a price for that board??

tensorflow runtime comes installed. Caffe torch or theano you have to build. ros-perception in dusty-nv’s github has full instructions for building the system. Digits you need an intel box with a gpu and its still going to take days to train a model. Or you can spin up an aws or ec2 vm and do it that way. mavros package gets you the interface to pixhawk within ROS. ROS has way more capability than dronekit. Most stuff you can just grab a package and build it into your system without writing any code. Pi’s and Odroids just don’t have enough cpu or gpu to handle DNN. tiny dnn or darknet is their limit.