Why it’s misleading? The driver was there to intervene when something was out of control. It’s not a perfect autopilot system yet. Why did they have the driver on the car? The woman was not the only victim. The driver could have been killed too if some heavy object suddenly blocked the road.
Seriously?!
https://www.seattletimes.com/business/apxreport-uber-driver-streaming-the-voice-just-before-crash/
Turns out, focusing on the driver was not misleading.
OK but it wouldn’t be unusual fora human driver not to pay attention to the road.
This was not a little distracted. This was totally disengaged. It would be unusual for a human driver to watch a show while driving. Which means the person who was supposed to keep eyes on the road was not doing her job.
Isn’t it what self-driving is for eventually? Why didn’t the car stop automatically?
Perhaps eventually, @igloo, but the technology may not be quite there yet. And there are issues like detour signs or people redirecting traffic with flags and such that need to be addressed…
what are the protections against hacking or computer failure? While that can happen to a human (sudden heart attack or other issue), it seems like a driverless car is vulnerable to failure. Look at what has happened with airline systems that bring air traffic to a grinding halt. How can driverless cars ever be totally hack- and failure-proof?
I guess not. Since they were doing the road test, I expected it more advanced than that. You don’t need a road test to know you need to be prepared for unexpected pedestrians on the road. Recognizing a pedestrian crossing is a low bar. Road tests should be reserved for things we can’t easily imagine.
We all drove with our teens. At some point, when they were not 100% ready for the real road, they had to get out on real roads with their learners’ permits and an adult in the seat next to them. Can’t imagine driving with my pre-DL teen and watching Hulu!
You couldn’t watch if you were allowed. Teens’ driving is nerve wrecking. This is different. It could be easily simulated without a road test.
Yes, it is not much different than teaching a teen to drive. No, it can’t be “simulated without a road test.”
Oh yes. Easily simulated. Have someone cross in front of the car and see it recognize it. Yo don’t need to be on the road for that.
Many of the comments focus on the technology having to be perfect to make them safer. It seems to me that driverless vehicles would become safer just by having more of them on the road. The more driverless vehicles, the fewer crazy decisions on the road by aggressive or impaired drivers. I am really looking forward to it.
I am curious as to the impact of having most road traffic be driverless technology. For instance, it seems to me one casualty might be short distance airplane routes, and some rental car business. Right now we love 45 minutes from the airport. If we leave an hour in advance, and need to get there 2 hours before the flight, then rent a car to drive to our hotel after a 2-hr flight, I’d really think three times about flying.
Recognizing a pedestrian on the road isn’t asking for perfection. It is so basic and should be easy. That they haven’t master it indicates how primitive the technology still is. Cynics in me hink they started road test not because of necessity but to increase visibility, nothing more than advertisement.
Its a lot more complex than that, @igloo, and no its not for advertisement- they have to get permission to try the vehicles in vivo. The software and cameras address size, speed, direction, predictive behaviors, etc.While its not perfect yet, its not primitive. And there are way more driver-caused accidents than driverless (even comparing percentages of vehicles to accidents). Take a look at the pdf from Waymo I posted. It provides a lot of helpful information on the many safety variables it involves, road hazards, behavior prediction, etc. Here is the pdf https://storage.googleapis.com/sdc-prod/v1/safety-report/Safety%20Report%202018.pdf They mention:
It timed out before I had a chance to mention- Waymo (that is Google’s driverless car) has been on the road since 2009. Primitive? Hardly.
Woz does not think driverless carriages will take over the roads any time soon.