RPi RP2040-based impulse must have CLI start?


Will a deployed impulse run without the ‘edge-impulse-run-impulse’ CLI command? It does not appear to. Everything works great with CLI init/USB board connection, a restart shows nothing happening without. I just want to confirm this is the case, do I need to write a ‘boot wrapper’ or similar construct?

Project ID:

Context/Use case:
Deployed firmware with integrated impulse (C++ library).

Hi Mike, thanks for your important question!
Make sure the call to ei_start_impulse(false, false) is placed correctly and not commented out.

Hi @moverturf,

The binary firmwares generated by the studio use AT Commands to start/stop inferencing. This can be done through our CLI or by sending AT commands through serial. To get the impulse to start automatically, you can recompile our base firmwares available on our github.