We’ve deployed several Jetson Nano SoM modules in our industrial setup, and we’ve noticed that some of them have stopped working after around 2 years of operation. Unfortunately, we don’t have a clear idea of how to begin troubleshooting the root cause.
Here’s what we know:
A serial console log is almost always the start. When you say there are no signs of activity, was this from a serial console, or something else?
If the units are on a custom carrier board, then probably the next step is to try them on a developer kit carrier board. Changes to device tree will mean that parts of the dev kit carrier won’t work if your device tree changes for your carrier, but the serial console boot log would still be useful.
If you know of any package updates occurring immediately before this, then that is worth mentioning.
You can clone the root filesystem and loopback mount this on a host PC to examine things like the last logs when it did work.
A clone can be applied to a new unit which is otherwise flashed for the same carrier board you expect. If the cloned rootfs works with a new non-rootfs content, then that can point out other software issues.