#16
|
||||
|
||||
The devil is in the details. The catch is in what Musk didn't say Having systems that can leave a parking space and come get you is the easy part. Doing so with an extremely low probability of running someone over is the hard part. The latter won't happen for quite some time.
__________________
Old'n'Slow |
#17
|
|||
|
|||
maybe they should keep the cars from running straight into a concrete wall at high speed first. Then we'll talk about moving forward with a more ambitious plan
|
#18
|
|||
|
|||
What happens if there is a software bug and driver is busy writing an email? System crashes, is there 100% failaover? How to driverless cars handle windy back road of New England, with ice and snow?
This is so scary (to me) |
#19
|
|||
|
|||
Quote:
|
#20
|
|||
|
|||
Quote:
15 year old, beat up trucks work fine, 15 year old self driving electric trucks ? |
#21
|
||||
|
||||
Can anyone explain how an autonomous car can position itself on a snow covered road?....or a dirt road?.....or new pavement that hasn't had lines painted yet?
dave |
#22
|
|||
|
|||
That's funny so true. Ok so it hasn't been a total bed or rose's with the 10 year old Subie. The airbag light came on and was bugging me much like the tps light that's been on since day one I got it last Aug and not going to fix the latter because it's called a tire pressure gauge and common sense. However if there's a problem chances are someone made a video on how to fix it and so I found and fixed...fingers crossed. The fix was reflowing/adding a bit of solder to a tiny set of resistors on a PCB connect to more silly warning lamps. Can you imagine something not going wrong with a self driving car?...lol.
|
#23
|
|||
|
|||
this is a really good question, and I think that's why any honest developer says it's too early. That was the problem that ran the tesla into the concrete barrier, they were relying on road lines and they were incomplete. The barrel covering the end of the wall was gone because of a previous crash. This kind of thing happens all the time on roads. Musk says they don't need lidar, i.e. it's too expensive, but cameras alone will kill people. Same thing driving into the side of the truck, a lidar would have been enough to keep that from happening.
|
#24
|
|||
|
|||
Have you priced a truck lately? TONS of the trucks you see on the roads in most urban-suburban areas cost at least that. |
#25
|
|||
|
|||
I guessing that what they say for marketing purposes is one thing, and the reality in the fine print of the user's manual for non-optimal conditions will be something completely different.
|
#26
|
||||
|
||||
The technology industry has a somewhat (in)famous track record of using end users as beta testers.
I hope self driving cars doesn’t become another example. It’s already unsettling (for me, at least) to see car commercials where manufacturers tout the safety features on their cars while showing a driver experimenting with taking their hands off the wheel. |
#27
|
|||
|
|||
And in a completely unrelated story my work computer died on me in the middle of entering spreadsheet data today. Good thing I wasn't behind the wheel when this happened. LOL!!
|
#28
|
|||
|
|||
Quote:
As of Dec. 2018, Waymo (Google) is "commercially" running Level 4 Autonomous vehicles in Phoenix -- geofenced (7x7 miles) and under fair weather conditions -- but, to the best of my knowledge, still has two engineers sitting in each car. Someone might have Level 3 in production, but I am not sure. |
#29
|
||||
|
||||
computer-operated vehicles are probably only 95% competent at avoiding collisions...
which makes them significantly safer than human-operated ones. most people absolutely suck at driving... e.g., if it's used at all, the turn signal goes on BEFORE the brakes are depressed and the steering wheel is turned... |
#30
|
|||
|
|||
Quote:
|
|
|