There are many safety features that go into luxury cars only, such as adaptive headlights, because I assume the mechanics and intelligence that go with them are too expensive to put in a budget vehicle and still make a satisfactory profit on them. I can't see automation being cheaper than something like adaptive headlights.
If they were sitting on the trailer with their legs dangling off and they didn't see the car in time they could become pinned or worse. Again, it doesn't really matter what the scenario is. A 4500 lb vehicle shouldn't be moving anywhere unattended, ever. When it does move on its own it shouldn't have a partial field of view with the sensors, that just leaves an opening to cause damage to the car. It should at minimum sense the entire front of he car, not just things that are low. To me this seems to be common sense.
If you are driving the vehicle, yes. But in an automated car, you are just a passenger and the computer is the driver. A passenger is not liable for how a car drives.
It depends on what insurance companies do. People won't be held liable for a mistake the automation may make. If I can insure the automated vehicle just as property and not have to pay the same liability for the question of my driving skill then that would definitely be a step forward. Perhaps car insurance companies will just be for manual cars and the automated vehicle will be rolled into house insurance. That would be fine. Also I think it not only depends on whether they are safer then a human or not, but the type of accidents they get into. If they are doing things that are just stupid like driving into a trailer when unattended, people are less likely to trust them and they are less likely to have adoption. I don't care whether the guy pressed the button or not, a 4500 lb vehicle should not be rolling anywhere unless all collision scenarios are detected.
So you're saying things are dangerous in the world so we shouldn't prevent more knowingly dangerous things from entering the world? That's pretty ridiculous. Listen, a 4500 lb machine is moving out of control here. I don't have to care if there may be an edge case that might be able to involve injuring a human. I'm not in a position to anyway. We now know of one blind spot for the sensors, this makes me wonder about other blind spots. Can the car differentiate between a child who has fallen or a homeless person on the road and a speed bump? Only the engineers know. This is why Tesla should be obligated to test every single scenario that can happen in the real world which obviously they haven't. They need to go back to the drawing board on this one. They have overstepped in their exuberance to release automation.
Wow you know what. I've been in a conversation about sensors before and people made them sound like they were absolutely perfect and could always see everything better than a human. Thank you for educating me otherwise.
Why because we won't have capitalism any more and self-driving taxi companies will no longer be looking to maximize profits? Or perhaps you will just wake up by then.
Yes because something which isn't perfect for use in every scenario shouldn't ever be released despite it having clear utility and working just fine in countless other scenarios.
What scenario do you have in mind that has enough practical value to make it forgivable that the care could kill or injure someone in the driver's absence? I'm really interested. Does it feed starving babies in Africa?
Companies like Tesla want all the advantages of selling an automated vehicle while accepting none of the responsibilities. Tesla has added a feature to move the vehicle automatically. They have a responsibility to ensure this feature can never, ever cause damage. They failed.
I think it is likely that the driver fucked up. I just don't believe that Tesla is free of guilt here in any case. A car should not be able to drive into things on its own, PERIOD.
Even if the driver did inadvertently activate the summon mode, a car that is smart enough to drive itself should also be smart enough to prevent a collision while doing so in *all situations*.
What if there was a person on the trailer? Seriously, this is just a bad design of the car and a demonstration of not having enough fail safes. You can't say it was an edge case like it excuses anything.
Car companies aren't adding the feature of 'roll when neutral' to the car, that is just a quality of manual transmissions. On the other hand, Tesla added the summon feature to their car so they should be making damn sure it is foolproof.
What the driver did do or did not do is totally irreverent. The car drove into something on its own. Automation failed. Recall all cars with this technology immediately and go back to the drawing board.
Personally I don't care if the user was lying, for a car to have the capability to do this is dangerous, reckless, and stupid. The user could come out and say, "yeah I did it as a test of the car's intelligence" and I would still side with the user because this is just so stupid. Non automated cars have fail safes, so should automated cars.
Slashdot Mirror
There are many safety features that go into luxury cars only, such as adaptive headlights, because I assume the mechanics and intelligence that go with them are too expensive to put in a budget vehicle and still make a satisfactory profit on them. I can't see automation being cheaper than something like adaptive headlights.
So what happens when there is no one in the car?
These days it seems they're willing to put anyone in any position as long as they will work for bottom dollar.... and it's not their position.
You traveled to France and ate at McDonalds??
If they were sitting on the trailer with their legs dangling off and they didn't see the car in time they could become pinned or worse. Again, it doesn't really matter what the scenario is. A 4500 lb vehicle shouldn't be moving anywhere unattended, ever. When it does move on its own it shouldn't have a partial field of view with the sensors, that just leaves an opening to cause damage to the car. It should at minimum sense the entire front of he car, not just things that are low. To me this seems to be common sense.
If you are driving the vehicle, yes. But in an automated car, you are just a passenger and the computer is the driver. A passenger is not liable for how a car drives.
You're saying automation is no more expensive then a catalytic converter? If that's the case then yes I guess it will be fine.
It depends on what insurance companies do. People won't be held liable for a mistake the automation may make. If I can insure the automated vehicle just as property and not have to pay the same liability for the question of my driving skill then that would definitely be a step forward. Perhaps car insurance companies will just be for manual cars and the automated vehicle will be rolled into house insurance. That would be fine. Also I think it not only depends on whether they are safer then a human or not, but the type of accidents they get into. If they are doing things that are just stupid like driving into a trailer when unattended, people are less likely to trust them and they are less likely to have adoption. I don't care whether the guy pressed the button or not, a 4500 lb vehicle should not be rolling anywhere unless all collision scenarios are detected.
So you're saying things are dangerous in the world so we shouldn't prevent more knowingly dangerous things from entering the world? That's pretty ridiculous. Listen, a 4500 lb machine is moving out of control here. I don't have to care if there may be an edge case that might be able to involve injuring a human. I'm not in a position to anyway. We now know of one blind spot for the sensors, this makes me wonder about other blind spots. Can the car differentiate between a child who has fallen or a homeless person on the road and a speed bump? Only the engineers know. This is why Tesla should be obligated to test every single scenario that can happen in the real world which obviously they haven't. They need to go back to the drawing board on this one. They have overstepped in their exuberance to release automation.
Wow you know what. I've been in a conversation about sensors before and people made them sound like they were absolutely perfect and could always see everything better than a human. Thank you for educating me otherwise.
Time will prove me right.
Why because we won't have capitalism any more and self-driving taxi companies will no longer be looking to maximize profits? Or perhaps you will just wake up by then.
Yes because something which isn't perfect for use in every scenario shouldn't ever be released despite it having clear utility and working just fine in countless other scenarios.
What scenario do you have in mind that has enough practical value to make it forgivable that the care could kill or injure someone in the driver's absence? I'm really interested. Does it feed starving babies in Africa?
Companies like Tesla want all the advantages of selling an automated vehicle while accepting none of the responsibilities. Tesla has added a feature to move the vehicle automatically. They have a responsibility to ensure this feature can never, ever cause damage. They failed.
I think it is likely that the driver fucked up. I just don't believe that Tesla is free of guilt here in any case. A car should not be able to drive into things on its own, PERIOD.
Even if the driver did inadvertently activate the summon mode, a car that is smart enough to drive itself should also be smart enough to prevent a collision while doing so in *all situations*.
What if there was a person on the trailer? Seriously, this is just a bad design of the car and a demonstration of not having enough fail safes. You can't say it was an edge case like it excuses anything.
So they even admit in the manual that the car is not ready for real life and the public. What idiot would buy something like this?
Just because it's an edge case doesn't excuse the fact that the car fucked up royally on its own.
If a car is sold as being smart enough to drive itself, then it should also be smart enough to prevent real damage from a human error.
Car companies aren't adding the feature of 'roll when neutral' to the car, that is just a quality of manual transmissions. On the other hand, Tesla added the summon feature to their car so they should be making damn sure it is foolproof.
I felt a second citation was needed: https://news.slashdot.org/stor...
What the driver did do or did not do is totally irreverent. The car drove into something on its own. Automation failed. Recall all cars with this technology immediately and go back to the drawing board.
The driver's alibi is totally irreverent. Cars should not drive into things on their own no matter what the driver does.
Personally I don't care if the user was lying, for a car to have the capability to do this is dangerous, reckless, and stupid. The user could come out and say, "yeah I did it as a test of the car's intelligence" and I would still side with the user because this is just so stupid. Non automated cars have fail safes, so should automated cars.
Apparently summon mode defeats the purpose of not having the car roll into anyone or anything.