driverless cars

[Randall Twede]

I don't think I'm a luddite. I have no problem with driverless cars. as long as they have a manual override. I wouldn't own one that didn't. furthermore I think it should be mandatory that they all have that. otherwise if your car hits someone, who pays? not you because the car was driving. not the manufacturer, you can be sure of that! but if they all have a manual override the driver/user/owner could be held responsible. I don't like computer chips in cars period. I heard a story about how someone hacked a car and took over some of its subsystems. not only that, but an EMP can knock out a cars computer. and having a chip in a car makes them much harder to repair. oh well, perhaps I am a luddite afterall

[Paul Clapham]

If I'm riding in an elevator, and it goes out of control and hurts someone, who pays? I don't want an implementation where the elevator has an override so that the rider can control it, for the purpose of making the rider (me) responsible for the result of crashes.

Because that's the only possible purpose of a manual override. Am I going to sit there at high alert for four hours just in case the driverless car I'm in has a problem? Not likely. For one thing I don't want to do that and for another thing it's unreasonable to expect anybody to do that. So, manual override just so I can be held responsible? Not for me.

[Randall Twede]

I see your point Paul, but I still wouldn't own one without that feature. what if the road is iced over and you hear on the radio that there is a wreck just around the corner. wouldn't you want the option of taking over?

[Paul Clapham]

The option of taking over? Sure, I suppose so. Maybe I want to drive over the sidewalk and through the bushes if there's a major problem. But the legal responsibility of taking over? Definitely not.

Anyway if the road is icy then the automated car should already be driving slowly enough that coming upon a wreck shouldn't be a concern.

[Jeanne Boyarsky]

Randall Twede wrote:otherwise if your car hits someone, who pays? not you because the car was driving. not the manufacturer, you can be sure of that!  

What about the insurance company?

[Ron McLeod]

Randall Twede wrote:.. what if the road is iced over and you hear on the radio that there is a wreck just around the corner. wouldn't you want the option of taking over?

We are not far off from having vehicle-to-vehicle and vehicle-to-infrastructure communications  (starting as soon as the end of 2017).  Your car will know that there are problems ahead will be able to take the appropriate action on its own - slow-down, take an alternate route, etc.

[Campbell Ritchie]

Randall Twede wrote:. . . what if the road is iced over and you hear on the radio that there is a wreck just around the corner. . . .

Don't talk to me about ice; my fingers still hurt from scraping the windscreen this morning. It took the best part of half an hour to find the glass so I could see whether the road was icy.
Surely a driverless car would slow to a sensible speed on ice, and would take the bend slowly enough to avoid the obstruction. Any panic is more likely to upset the control of the vehicle and increase the risk of a collision. And what about what I saw yesterday, somebody falling off his bicycle because Diamond Street was all icy. The driverless car would have to be programmed to avoid him too. Which is more than human drivers would be, since he was wearing the local uniform of black from head to toe and no lights.

Declaration of interest: I was on two wheels too, wearing yellow, with lights. I checked he was all right and told him to walk to Union Street and ride there so as to avoid the ice. It is a long time since I last fell over on ice.

[fred rosenberger]

Jeanne Boyarsky wrote:What about the insurance company?

Who's insurance company - mine as the owner? the manufacturer surely has insurance. The victim has insurance. The manufacturer has the deepest pockets, so if there is a lawsuit, guess where everyone is going to be reaching?

[Paul Clapham]

From what I've read (and sorry I can't provide any links), most manufacturers of driverless cars are skipping the part of the process where the occupant of the vehicle has the option of controlling it manually. I think that's because of the issues raised here: in practice the occupant isn't guaranteed to take control when crap starts to happen, and if the occupant does take control they aren't guaranteed to do anything helpful. And then there's the liability issue: who gets to take responsibility for the result? So that's why the manufacturers are fine with features like automated parallel parking, and some have tried partial automation with occupant override but there's been at least one death there and quite a few near-misses. The near-misses can be kept quiet, so we don't get to hear much about them, but deaths can't.

[salvin francis]

It's all in the news here about a [some huge cab company name here] driverless car running through a red signal while a pedestrian was crossing the road. I wonder what would happen if a manual override was provided.

[Bear Bibeault]

salvin francis wrote:It's all in the news here about a [some huge cab company name here] driverless car running through a red signal while a pedestrian was crossing the road. I wonder what would happen if a manual override was provided.

By the time the occupant realizes and reacts, it's likely too late.

[Maneesh Godbole]

salvin francis wrote:I wonder what would happen if a manual override was provided.

The defence lawyers would be most grateful I suppose. If the human hadn't taken over, none of this would have happened. Blame the human. Our cars are perfect.

[Maneesh Godbole]

Quite some valid and interesting points have been raised above.
I think the correct way to go would be dedicated roads/tracks where ONLY driverless cars are allowed. No human intervention at all. Enclosed environment with all car computers in constants communication with each other would be as close to "fool proof" as we can get.

[Jeanne Boyarsky]

Maneesh Godbole wrote:Quite some valid and interesting points have been raised above.
I think the correct way to go would be dedicated roads/tracks where ONLY driverless cars are allowed. No human intervention at all. Enclosed environment with all car computers in constants communication with each other would be as close to "fool proof" as we can get.

I think that's the future. The hard part is getting there from where we are today.

[Randall Twede]

I think what will happen is we will still have to buy liability insurance, but it will cost much less because the self-driving cars get in less wrecks.

[Liutauras Vilda]

Jeanne Boyarsky wrote:

Maneesh Godbole wrote:Quite some valid and interesting points have been raised above.
I think the correct way to go would be dedicated roads/tracks where ONLY driverless cars are allowed. No human intervention at all. Enclosed environment with all car computers in constants communication with each other would be as close to "fool proof" as we can get.

I think that's the future. The hard part is getting there from where we are today.

That isn't a future anymore, that is something already exists. London's Heatrow Airport Pods.

[Paul Clapham]

A related topic: Nissan is proposing automated cars which call Customer Support when they encounter a situation which they can't deal with.

Article here: Nissan uses NASA rover tech to remotely oversee autonomous car.

There's a picture there of a situation which an automated car might not be able to deal with, in which case it would call for help. This is where I started laughing.

Automated car: Hello, I have a situation outside of my parameters.

Support: Your call is important to us. Please stay on the line as all of our agents are currently busy and calls are handled in sequence. Your estimated hold time is... forty... three... minutes.

[Campbell Ritchie]

Paul Clapham wrote:. . . Your estimated hold time is... forty... three... minutes.

It said 90″ on that link. A 90″ delay at roadworks would be enough time for the following driver to string you up from the nearest lamp‑post round here, so pray nobody nearby has any rope handy

[Paul Clapham]

Campbell Ritchie wrote:

Paul Clapham wrote:. . . Your estimated hold time is... forty... three... minutes.

It said 90″ on that link.

Yeah, it might be 90 seconds now but they are still trying to sell the technology. Wait until people start using it and the cost-cutting begins. Then you'll see the wait time go up.

[Randy Maddocks]

I am on the fence about this kind of technology. On the one hand I am all for technology that can potentially help avoid accidents when a human fails to sense impending danger (e.g. blindly changing lanes and not seeing a vehicle in the lane the driver is attempting to change into). On the other hand, a driverless car worries me. I think most would agree we humans like to have a sense of control, or at least know we can take necessary action to be able to take control of a situation. But to completely and wholly depend on a driverless car to navigate the highways and byways of the land, I am not sold on.

Tesla recently reported an accident with one of their driverless cars where the occupant of the vehicle died (article). Granted, one point stated in the article was that this was only one fatal accident of out millions of tests already done and that human error happens all the time on the roads. But the fact that the car failed to detect a tractor trailer in front of it does not give me the "warm and fuzzies" about being in this kind of car.

I also share Randall Twede's concerns with computer chips in cars. There is no doubt it is amazing technology, but the fact that someone could hack into the system and make it do who-knows-what while you are speeding down a highway, does not bode well with me.  It would appear I could also be considered a "luddite".    

[Christian Peacock]

I'm against driverless cars because the problems they set out to solve are mostly just symptoms of the real problems - there are too many people and too many people rely too heavily (through no personal fault of their own) on cars. Most car journeys now are through necessity and by automating the process with no opportunity for manual control, we are surrenduring to this fact and eliminating the opportunity for journies purely for pleasure's sake.

I do not think the human race should continue the pursuit of relinquishing control in the name of safety. There will come a time when everything we do is entirely without risk, or any kind of stimulation and we will then ask ourselves, "What is the point of being alive?"

[Campbell Ritchie]

Apart from the increased population, for the last fifty years at least there has been a tendency for people to live farther from where they work, requiring much more travelling. Some of that is by their own choice, some forced on them. About 1½ miles from where I live, the council compulsorily bought and demolished several hundred little terraced houses, forcing their occupants to move elsewhere, usually much farther from work.
There is a difference between travelling for pleasure or not, and removing all risk from life. As soon as you start taking risks whilst travelling, you are inflicting risks on others, which is doubtless wrong. Any risks taken to get one's adrenaline levels up shou‍ld be where innocent people and uninvolved third parties are not put at risk.

[Christian Peacock]

Campbell Ritchie wrote:Apart from the increased population, for the last fifty years at least there has been a tendency for people to live farther from where they work, requiring much more travelling. Some of that is by their own choice, some forced on them. About 1½ miles from where I live, the council compulsorily bought and demolished several hundred little terraced houses, forcing their occupants to move elsewhere, usually much farther from work.

Totally agree, this is exactly the kind of thing I'm talking about. When I started as a software developer in 1997, I worked somewhere that was about 2.5 miles from where I lived, and often walked. From that place there was a choice of 2 places to go and buy food (not in a town) within 5 minutes walking distance. Now, all 3 of those premises have gone, two of them to be replaced by houses. I now drive over 30 miles each way to work.

Campbell Ritchie wrote:
There is a difference between travelling for pleasure or not, and removing all risk from life. As soon as you start taking risks whilst travelling, you are inflicting risks on others, which is doubtless wrong. Any risks taken to get one's adrenaline levels up shou‍ld be where innocent people and uninvolved third parties are not put at risk.

What I'm talking about with this is not basically the use of public roads as racetracks, but just the pleasure of being in control of a vehicle and the use of skills. I wonder, how many skills can the average person claim to have? And how many might that have been 30 years ago? Many skills are being diminished under the "leave it to the experts" (or machines) mentality and driving is one remaining skill posessed by many that really taxes the brain. We shouldn't be eager to lose the opportunity to have such skills - there is more to life than being safe.

[Campbell Ritchie]

I see what you mean about driving now