- This topic has 15 replies, 10 voices, and was last updated 8 years, 3 months ago by
.
-
Topic
-
-
Replies
-
Nobody will ever come out and say what failed and caused the accident. You can bet that Tesla’s lawyers either have or will settle with the family quickly and quietly.
Ultimately it will be an important step towards the first regulations for these systems. As far as I understand it, these systems are still unregulated themselves and having no way to measure the amount of “aid” they provide. There’s mainstream vehicles that are getting pretty damn close to being able to drive themselves in some circumstances with automatic braking, lane keep assist, adaptable cruise control etc. But as of right now, there’s no measurement for it.
At least one of Google’s self driving cars was even in an accident.
https://www.engadget.com/2016/02/29/google-self-driving-car-accident/
With a product that is potentially as lethal as an autonomous car, there has to be regulations of some sort. There’s no 2 ways about it. If not, who decides when a car is actually autonomous? And who would decide if an “autonomous” car can actually safely operate autonomously. Honestly I could call my SRT4 autonomous and there’s no measure to prove me wrong because opinions are still not fact.
First
Action
In
LearningThe question of insurance is interesting.
If a vehicle does something autonomously, which results in personal or property damage, who is at fault? Who pays restitution/compensation? You, the policy holder, were using the vehicle as the manufacturer intended, you weren’t negligent and the accident occurred without input from you.
I think self driving cars will be common in the future and the technology will be much better than the average human driver, but It seems the manufactures are relying on the human driver as the backup system.
Regulations are a tricky subject. The manufactures usually like them because if the vehicle meets a “regulation” that can claim the system is “good enough” to prevent potential lawsuits, and may result in a less expensive and robust system that might have been developed otherwise.Computers can perform complex calculations in the blink of an eye. But, humans have innate pattern recognition skills that cannot be matched by current technology. There was a study on how soldiers became adept at spotting IEDs. They found that experienced soldiers instinctively spot things that were out of place and areas did not have enough natural randomness. We are constantly scanning for signs of a “threat” while we are driving, even though we may not be consciously aware of it. It will be interesting to see when the algorithms match or exceed human skill and if humans will eventually “forget” how to drive.
It would appear that Tesla is on a streak.
http://jalopnik.com/driver-in-tesla-model-x-crash-cited-for-careless-drivin-1783476762
http://jalopnik.com/another-tesla-model-x-crash-blamed-on-autopilot-this-t-1783480364
They don’t have to be perfect, only have to beat their average human counterpart, which isn’t setting the bar very far out of reach.
What I don’t understand is why hasn’t the truck driver been held responsible? The Car had the right of way, and we have all seen truck drivers do exactly what happened here, where they pull out into traffic hoping everyone stops for them. The truck driver should never be allowed to drive a truck again, and should face some type of manslaughter charge.
– I am a CDL class A license holder, I drove the lower 48 for two years back in 2006-2008, and there is no excuse for the bahavior of many truck drivers. Most of them are rather stupid, and have no patients and because they are paid per mile and have deadlines, and they like to waste time, it causes them to be a bit anxious and take chances. No excuse for what the driver did.This is the part of technology that is really difficult to figure out. How do you assess blame when technology fails? What is the reasonable liability for everyone involved?
In my real life work, I write software for the RF side of cellular systems. This code needs to be highly reliable, and knowing the number of permutations I need to consider I’m amazed that the autonomous mode on cars works as well as it does. And how do they program the fail safe modes? Was this kind of issue foreseeable? Unfortunately, this will all be hashed out by lawyers. Al new technology has things like this happen. This technology is created by humans,and we all know humans are fallible.
I’m impressed with the technology that Musk’s companies produce. Spacex can land a rocket on a barge at sea. This incident with the Tesla would not stop me from considering one (although, since I like to drive, auto mode would not be enabled). BTW, if you get a chance to drive a Tesla, do it. The torque is amazing.
I wonder how many of Tesla’s self drive cars have crashed , compared to the crashes down to human error .
Sent from my iPhone using Tapatalk
[quote=”dtidman” post=170039]This is the part of technology that is really difficult to figure out. How do you assess blame when technology fails? What is the reasonable liability for everyone involved?
[/quote]It is hard to say how this will play out in a world where people expect perfection from everyone but themselves. In the end, the decision will be probably be made by our legislators on the Supreme Court.
I think this is one of the things that Kubrick was trying to say in 2001: A Space Odyssey. Beyond a certain point, relying too much on technology has its issues.
[quote=”Evil-i” post=170141]I think this is one of the things that Kubrick was trying to say in 2001: A Space Odyssey. Beyond a certain point, relying too much on technology has its issues.
That’s a very good point …..Dave
Sent from my iPhone using Tapatalk
[quote=”Commodore665″ post=170106]I wonder how many of Tesla’s self drive cars have crashed , compared to the crashes down to human error .
Sent from my iPhone using Tapatalk[/quote]
I think the sample size of self driving cars is still too small to compute the statistics?
On the tesla/truck accident, I read that the radar sensors were too low to sense the tall truck trailer? If that is true, then I think it would be a design flaw if they can no detect to at least the cars height to go under? If the sensor did detect the trailer, but the software ignored it, then there software had issues.
[quote=”Evil-i” post=170141]I think this is one of the things that Kubrick was trying to say in 2001: A Space Odyssey. Beyond a certain point, relying too much on technology has its issues.
Nailed it!
[quote=”Evil-i” post=169964]The question of insurance is interesting.
If a vehicle does something autonomously, which results in personal or property damage, who is at fault? Who pays restitution/compensation? You, the policy holder, were using the vehicle as the manufacturer intended, you weren’t negligent and the accident occurred without input from you.[/quote]Since the system is not truly autonomous, the person behind the wheel is still the driver who is responsible. It’s a misnomer to call it auto-pilot. Treat it as advanced cruise control, because it is not intended for the driver to stop watching the road and fully trust the car. If the system is having trouble, it will tell the driver to take over. Case in point, an Asian man was driving a Tesla along a winding, mountain road. He turned on the “auto-pilot” and sat back. The car tried getting him to take over, however he didn’t understand English enough to understand the warning, and the car crashed. Sure there was a language barrier, but the car was alerting to it’s inability to handle the road.
I use cruise control, I still check the speed occasionally. Besides, there are already laws about distracted driving. Just watching a movie behind the wheel means he was breaking the law.
- You must be logged in to reply to this topic.