The sensor range rules are confusing. SR4A 334 says that "each [sensor] package has a sensor range that indicates the limits of the sensor's reach (see the Signal Rating Table, p. 222)". I presume this means that unless the description of the individual types of sensor states a specific range, I compare the sensor's rating to the Signal Rating Table and that gives me the range.
But when I do that, I'm surprised to discover that the the rating 1 sensors on most civilian vehicles can't detect anything more than 40 meters away - not even the cameras, by a strict interpretation of the rules. One wonders how the Pilot can maneuver through traffic at freeway speeds of 90 meters per turn or more when it can't detect brick walls at a range of 41 meters.
But the individual entries on the Signal Rating Table offer examples of the kinds of sensors typical of each rating, and vehicle autonav sensors are listed at rating 3, with a range of 400 meters, which is more useful. However, that still says that a metahuman observing the world through a set of vehicle sensors will not be able to see as far as he could with his Mark I Eyeball. With sensor performance as poor as that, who would want to leave the driving up to the car?
If I presume to use the Signal rating of the vehicle to determine sensor range, I get a useful range limit, but I still run into the problem that the vehicle can't detect Mount Rainier at 401 m, which, at 90 meters every three seconds, is less than fifteen seconds down the road. So why would anyone let their vehicle drive itself?