I don't believe it's a critical component of Autopilot. From the last public video release in 2019 [1], it seems like they use the stereo cameras to derive depth information, and use radar as a kind of ground-truth in informing the model. I would think the source-of-truth for what the Autopilot system thinks is the position of objects is the stereo camera, i.e. the radar is not part of the online flow, but is used offline to correct the model.

See specifically at 2:20:22 [2]:

> We use that radar to annotate what the vision is seeing, the bounding boxes that come out of the neural networks, so instead of human annotators telling you that this car, in this bounding box, is roughly 25 meters away, you can annotate that data much better using sensors.

I take that to mean that the radar is used as part of supervised learning for updating the model offline, whereas just the cameras are used as input to the model.

[1] https://youtu.be/Ucp0TTmvqOE?t=8254

[2] https://youtu.be/Ucp0TTmvqOE?t=8422