Driverless Cars

[cqaigy2]

Autopilot was on, in that Mountain View crash.

Tesla: Autopilot was engaged in fatal Model X crash

That the exact problem had been noticed before, and the owner still engaged autopilot later is a good example of why I started this thread........ even people who should know better put too much faith in technology.....

Rgds, D.

"noting that a crash attenuator safety barrier intended to mitigate the effect of a collision in that spot was missing due to an earlier crash" I wonder if he had the autopilot going during the prior crash. And then there's person three cars ahead of me going to work the other day swerving across the center line repeatedly. I'm guessing a phone was involved.

 

[JerryK]

I don't think we will see actual year round/all weather driverless cars soon. I don't see anyone out here in the boonies want to drive the speed limit. And once a few cars pass them, they will sell it and get something faster. And, a lot of us can't wait to F with them and see what happens on the road. Probably be a ' F with the driverless car ' movement....

 

[k0ua]

My analysis of the accident from watching the very limited video, is that the accident is mostly the fault of the person crossing the road walking the bicycle, with her head turned totally opposite to traffic looking down at the bike. She was walking on the wrong side of the bike with her back to traffic. Now that said, perhaps, a very attentive human driver with the lights on BRIGHT, not dim, and not looking down at his cellphone MIGHT have saved her life. MAYBE, not certain. With the lights on dim and no reflective clothing or anything on the human or bike, even with the lights on bright, I am not sure she would have been seen in time. Why the robot driver didn't see her in time, I don't know. Because I don't know anything about the sensors and what their capability's are.

All I know is that an attentive human driver would have had a hard time preventing this accident, especially with the lights on dim. As far as the guy behind the wheel, well he was just a passenger, as all of them will be after a few minutes of being lulled into the sense of security and boredom that the robot driver provides.
There doesn't seem to be a reason to even have the human "backup" because he will always be bored out of his gourd, and is in no way able to see and prevent an accident. He might be useful if the car breaks down to call a tow truck or something, but I am sure the car can call for its own assistance as well.

 

[3930dave]

"noting that a crash attenuator safety barrier intended to mitigate the effect of a collision in that spot was missing due to an earlier crash" I wonder if he had the autopilot going during the prior crash. And then there's person three cars ahead of me going to work the other day swerving across the center line repeatedly. I'm guessing a phone was involved.

From what I've read on the Model X Mountain View crash/fatality, the vehicle had not crashed previously. BUT, what his brother was saying was that the owner had noticed the autopilot swerving towards that exact barrier previously.

I'd like to think that autopilot on these has to be manually activated, but I don't know any Tesla owners, and TBN is mostly the wrong demographic to ask..... Should be Opt-In for use, but :confused3:

It sounds like that particular safety barrier had been hit earlier by another vehicle, and Tesla is now trying to lay off some of the blame on CalTrans (or whoever else is responsible for that stretch of road) for not repairing the barrier fast enough.

Even end-on, if these systems can't detect multi-ton concrete barriers in their path then they definitely are not ready for prime time.....

Rgds, D.

 

[rd_macgregor]

The more of this I see the more convinced I am that the insurance industry could be the biggest impediment to widespread adoption of the technology. If I were an insurer, I'd want a HUGE premium markup to cover the uncertainties. Worse (for drivers/owners), every driverless car system says the human operator must remain vigilant at all times to override the system should it become necessary...I'm betting insurers will likely require cameras inside and outside the car so they can see whether to deny claims if the human is distracted when a crash occurs. Determining who is "at-fault" will be complicated, so a lot of lawyers will be fully-employed.

 

[3930dave]

Determining who is "at-fault" will be complicated, so a lot of lawyers will be fully-employed.

In my lifetime, I've watched govt and lawyer count expand dramatically..... and it will probably continue that way....... but, even the insurance industry is scratching their heads over the future....

Setting precedence will keep lawyers in Benzes in the near-term..... crystal ball gazing further out is interesting though..... if the major liability resides with the control software cut by a few companies, then it mostly will come down to a handful of car companies battling in court with a handful of insurance companies..... That scenario may thin the lawyer herd down, if it comes to pass.....

While I'm not a fan of bubble-wrapping everything (it doesn't solve all problems, and arguably, creates others....), I'd say there is a point about the barrier repair taking too long @ 11 days......

ONLY ON ABC7NEWS.COM: I-Team investigates why CalTrans didn't fix safety barrier before Tesla driver died there | abc7news.com

Rgds, D.

 

[Oaktree]

The more of this I see the more convinced I am that the insurance industry could be the biggest impediment to widespread adoption of the technology. If I were an insurer, I'd want a HUGE premium markup to cover the uncertainties. Worse (for drivers/owners), every driverless car system says the human operator must remain vigilant at all times to override the system should it become necessary...I'm betting insurers will likely require cameras inside and outside the car so they can see whether to deny claims if the human is distracted when a crash occurs. Determining who is "at-fault" will be complicated, so a lot of lawyers will be fully-employed.

So if the operator needs to remain vigilant and be able to take over at an instant, what is the purpose of a "driverless" car in the first place?

 

[Jstpssng]

So if the operator needs to remain vigilant and be able to take over at an instant, what is the purpose of a "driverless" car in the first place?

Too many out there seem to have problems remaining vigilant anyways, and it's been that way since long before cell phones and driverless cars.

 

[k0ua]

So if the operator needs to remain vigilant and be able to take over at an instant, what is the purpose of a "driverless" car in the first place?

I agree, until the "driverless car" is truly driverless, then they will be a hazard. Like all new tech, I am sure it will happen someday, but we are not there quite yet. No one can sit in a car with absolutely nothing to do and remain truly vigilint. Not for long anyway. Your attention will wander with the oncoming boredom as the car runs you and itself into something. People that think you can have a backup driver, have no understanding of people. Or driverless cars for that matter.

 

[rd_macgregor]

Driverless cars might also (eventually) lead to changes in DUI law enforcement. If the human is legally over the alcohol limit, but the car is driving itself, you might plausibly argue that the vehicle is safer...but is the inebriated "driver" still arrestable? I'll bet so...but a couple of decades from now, you might have young kids going places in driverless cars (taxis?), so why not drunks?