I wonder to what extent they can use the ultrasonics to validate the camera based blind spot recognition. Can they use them to measure the false negatives from the cameras to gain confidence before deploying EAP auto lane change?
At highway speeds, the ultrasonics are pretty pointless for BSD. They are just too slow.
For example, in the FSD promo video, there is a shot where a motorbike overtakes the car. In that scenario, ultrasonic BSD would not show the bike until it was out of the blind spot and alongside the front wing...
IMO, use them for avoidance, which the seem very good at, and use the cams for BSD.