The AI deck enables WiFi communication with the Crazyflie® as well as using the power-efficient GAP8 to run neural networks on board. This tutorial gets your AI deck to stream camera images to your computer with NanoCockpit, which we recommend for the AI-deck. Each step links to the documentation in our GAP8 example repository, where you find the details.
Make sure that you have gone through the Getting started with the Crazyflie 2.0 or Crazyflie 2.1(+) tutorial first.
To run this tutorial and set up the AI deck you will need the following:
The AI-deck can be mounted both on top and on the bottom using the provided long header pins. Either position might yield different pros & cons but for development when a JTAG cable needs to be attached the top position works best. Note though that the ESP32 antenna is a bit fragile and can be damaged during crashes. For better resilience during crashes, mount the AI-deck at the bottom, or put a peace of hot-glue as shown in the picture below.

NanoCockpit needs its own firmware on both chips of the AI-deck: the ESP32 firmware for NanoCockpit, and a NanoCockpit application on the GAP8.
With the AI-deck mounted, update the Crazyflie to the latest release, as described in the firmware upgrade section of the cfclient user guide. The release also contains the ESP32 firmware for NanoCockpit, which is flashed to the AI-deck automatically.
When it runs, the AI-deck creates the WiFi network aideck-nanocockpit.
Flash the NanoCockpit streamer to the GAP8 with the JTAG programmer. It is released as a ready-made image, so there is nothing to build, see flashing the example.
The viewer runs from a copy of the GAP8 example repository. If you do not have one yet, clone it:
git clone https://github.com/bitcraze/aideck-gap8-examples.git
Then install the viewer, connect your computer to the aideck-nanocockpit network and start the
viewer, as described in
starting the viewer.
The viewer shows the camera images of the AI-deck.
Your AI deck is now running NanoCockpit. From here you can
The other examples in the repository use the original ESP32 firmware instead. See switching firmware for how to go from one to the other.