Best Practices to Interface Isaac with Jetson GPIO

Hi there,

I was wondering about the best practices of interfacing GPIO scripts with the Isaac engine? The Jeston Nano OS image comes with the Jeston-GPIO in /opt/nvidia/jetson-gpio, but these are all python scripts. Should we somehow make calls to an external python script from C++ scripts in the engine? Or is there a reliable C++ GPIO library that can be integrated directly into the Isaac build? Some insight would be very much appreciated.



Any thoughts on this? On the one hand you have a great Jetson.GPIO library. On the other you have the Isaac engine. How do you integrate the two?


Do you mind telling what you plan to use the GPIO for?

Hope this helps -->

Hi Shrinv,

Thanks for getting back to me. I’d like to trigger pins on the board HI/LOW based on object detection in the Isaac engine. The Kaya robot example would move towards an object after detecting it. How did you guys accomplish that? Using an aux board like an Arduino doesn’t make sense given that the Jetson Nano has everything I need.


This may be helpful to you:

We will also check on how to best get from here to interfacing with Isaac and the Kaya app.

Thank you for your feedback.

Did you ever figure that out? Because this doesn’t make sense to me too and they’re literally all NVIDIA products. So it should make sense that NVIDIA Isaac SDK can communicate with the NVIDIA Jetson SDK’s GPIOs.

Can you provide a follow up on that please?

There is no built-in support for Jetson.GPIO yet, but you could write your own Isaac SDK Python codelet that uses Jetson.GPIO for your project.

1 Like