The system can sort of handle some of the hairpin turns, but it has a long way to go.
If you're not familiar, Lombard in San Francisco is one of the steepest and curviest streets in America. It's actually not easy for many seasoned drivers to deal with. However, since it has eight hairpin turns over a very short distance, you have to take it very slowly, so at least there's little chance for someone to be injured. However, it wouldn't take much to curb your car or hit a gate, railing, or landscaping.
Watching this Tesla Full Self-Driving Beta test reminds me of first teaching my daughter to drive. There were a few occasions I had to grab the steering wheel since she had a tendency to turn too soon or too late until she got used to the car. Like we said before, these cars should require signage to let people know they are "student drivers."
It also reminds me of driving behind drunk drivers. While we have no way of knowing for sure, we bet that any number of accidents happening on Lombard Street could be the result of someone having had too much to drink. Drivers behind this Tesla Model 3 may have been aware, or at least come to the conclusion, it was on Autopilot. However, if they weren't aware, it wouldn't be unfair to assume the "driver" was intoxicated.
This is not to say that the technology is completely unable to handle the twisty residential road, but it struggles on several of the hairpins. To be precise, it handles about three of them pretty well, and almost a fourth. In addition, it "maps out" the street and "sees" it just fine. In some cases, it appears to understand the tight curves and correct itself as it goes. However, if the driver hadn't helped, it wouldn't have made the tight turns safely.
Check out the video and let us know your take by leaving a comment below.