Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
Key word- "much less likely." This is really situational. Someone will have to put the understanding into the computer on how to react for every possible driving scenario ever. "The computer will know" is a dream answer not on par with reality. Traction control is a very limited scope assistance, as are most computer assist technologies in cars. I am not downplaying the amount of engineering that goes into them, quite the opposite - if traction control, which is a relatively simple task, is a complex engineering piece, then controlling the entire car is going to be several orders of magnitude above and beyond.
My next point is going to be on how the computer knows. The car will need a complex array of sensors ($$$) to even know whats going on. And how will the computer understand when one of those sensors is giving faulty data. Rain, snow, mud, etc are all going to have an affect on the reliability of any sensor that isn't purely vision based sitting in the car.
I'm sorry but Mercedes system doesn't impress me that much. Its a distance sensor on a car. Feed that data into a very simple calculation that gives you the rate of closure and pair it with an expected stopping distance given the vehicles current speed. I'm sure i'm oversimplifying the engineering that has gone into it, but at the end of the day its a small scope - when the rate of closure comes up on the threshold of expected stopping distance, apply the brakes.
And none of this still addresses the core argument - that a computer system WILL fail, people will get hurt, and the manufacturer of that computer will be liable for the damages.
Quote:
Originally Posted by sean98125
The car hits a puddle of water and hydroplanes, and in trying to keep the car on the road the computer over-corrects,
But that's just it - the computer is much less likely to overcorrect because it is doing only one thing - driving the car. It is taking in more pieces of information about road conditions, wheel speed and traction than any human ever could, and it is responding far more quickly than any human could. Most cars sold these days already have traction and stability control and ABS, which compensate for human errors. There are cars out there that prime brakes for rapid stopping when their sensors tell them that the driver is about to collide with something - the car is ready for it before the driver. (Look up Mercedes active braking system.)
Computers aren't drinking coffee or fiddling with the air controls. They aren't jabbering away on the phone or texting Brittany about what Jason said yesterday. They aren't putting on make up. They aren't thinking about being in Hawaii instead of in traffic. They aren't listening to talk radio and getting themselves worked up in a lather. They aren't taking risks in order to get to work a few seconds faster. They don't have egos to bruise. They are driving.
A computer could be programmed to notice a variation in another cars path that indicates hydroplaning and could immediately begin to prepare itself for it. If it senses that a wheel is spinning at a different rate than the others it could apply the brake on that particular wheel or all the others or any combination of wheels. (This is how stability control already works.) It could adjust the steering inputs incrementally to keep the vehicle stable. If there is a fishtail it could bring the car under control by applying brakes and steering input in a way that no human ever could. A human can't apply brakes on just one wheel, or 25% in one wheel and 80% on another. It's all or nothing for us.
When you slam on your ABS brakes, you are pretty much pressing a button that tells the computer "stop me as quickly and in as straight a line as possible." When you hit the gas in the snow, you are telling your computer "get me moving without skidding too much."
Automated highways will be the biggest safety improvement in transportation since the seatbelt and engineered crumple zones.
It is far cry from a perfect system and it's a far cry from a fully automated transportation system.
I wonder how motorcycles will be effected and tracked by this system?
Computers crash, fail and are hacked, I'll keep my life in my hands thanks. Hey, look at that huge pot hole.
As the driver is taken out of the equation they will not be paying attention and will not be able ti take over in time in the event of a accident at the elevated speed limits this system may allow.
Now it's winter and a blizzard come in dumps 2ft of snow and the viability is O, even the best Distronic Cruise or crash avoidance systems are going to be effected.
You go on about how folks don't know who to drive in such conditions. How will they now cope with this predicament of weather or a system failure as they will have very little experience driving a car as the "computer" has none all of it for them?
It is far cry from a perfect system and it's a far cry from a fully automated transportation system.
You are absolutely correct. It has taken about 17 years to get to this point with Distronic systems, with the first debuting on Mitsubishis in 1995. It's a great system and a key component to an "intelligent car", but a lot more is needed to make the concept of automated driving even in a highway envrionment a reality.
The European Commission launched an initiative in 2007 for advancing the "Intelligent Car" as part of the universal European i2010 initiatives. The main point of the program is to begin aiding in the development of systems to get to the intelligent car. Most of the work currently focuses on Distronic aka ACC systems, lane departure systems and driver drowsiness monitors. Their goal is to get 3% of cars on Europes road utilizing these technologies by 2020. The long term goal is to have a fully autonomous/intelligent car available for purchase in the 2040-2050 timeframe, that can operate on a road without needing any infrastucture changes or relying on every other vehicle around it to be intelligent.
Right now, we could have fully autonomous cars. GM and Google have both done it, but only on modified roads or in closed systems neither of which can be used without major infrastucture changes. Even the concept of the "highway as rail" where the car enters the road and then drives itself on the highway would require a massive effort to implant sensors and magnetic strips in the roadway as well as having cars that could utilize such systems. Even then, these systems don't react very well (at present) outside of predictable linear environments. The only way anyone has really gotten it to work is if the cars can "talk" to one another, but that is impossible to do in a country where the average car is 10 years old.
I think it will happen, I just don't think it will happen in 15 or even 20 years outside of self-contained systems like the one GM is currently testing.
No one is talking about it because its NOT going to happen.
its going to happen when taxi drivers start to loose their jobs. how is it not going to happen? if taxi companies start to buy these driverless taxis, what good is the driver for?
Its like those self checkout lanes at a grocery store, 1 person handling 6 checkout machines = 8 jobs lost, meaning 6 cashiers and 3 baggers would of been there, but there is just 1 person helping. sure, 6 cashiers would be a bit quicker, but you get the point, i hope.
You're telling me that in situation #2, you think the insurance of the driverless car is going to pay out and call it a day? You don't think its at all likely that:
1> injured people/insurance will sue the manufacturer of the computer driving the car for damages
2> person driving the car will sue the manufacturer of the computer driving the car for damages
Again, if you were in a bus and the bus DRIVER made a mistake and caused an accident, who is at fault? The bus driver, or you, who was only a passenger? Of course, the bus driver (and by extension his place of employment since this is a work situation) is liable. Now replace "bus driver" with computer.
You're right though, the liability hasn't changed. The "driver" of the vehicle is liable, which is now a computer (and by extension the computer manufacturer). Ergo, you'd have to be nuts to manufacturer autonomous cars for regular people, because then you'd be liable for all their crashes!
If that doesn't make any sense... good job trolling me! :-D
You can't sue a computer. If adding safety and convenience features were such a huge liability, car manufacturers would never do it. As in an autopilot-equipped plane, the pilot is still ultimately responsible for what the plane does. Wouldn't be substantially different for an automated car where control is still ultimately left up to the driver.
You are all thinking small ball on this car control thing. I don't care to debate it with small thinkers, but suffice it to say that a lot more complex problems than this have been easily solved when the time comes. For my part, I don't care how it works out since I'll be pushing daisys. But for you doubters, I can just see you questioning how the horses will be able to tolerate cars clogging up the roads.
Last edited by Wilson513; 01-26-2012 at 04:23 PM..
You are all thinking small ball on this car control thing. I don't care to debate it with small thinkers, but suffice it to say that a lot more complex problems than this have been easily solved when the time comes. For my part, I don't care how it works out since I'll be pushing daisys. But for you doubters, I can just see you questioning how the horses will be able to tolerate cars clogging up the roads.
Yea and in the 70's everyone thought we'd have flying cars by now. Like it or not the details, both technical and legal, need to be resolved before anything happens.
You can't sue a computer. If adding safety and convenience features were such a huge liability, car manufacturers would never do it. As in an autopilot-equipped plane, the pilot is still ultimately responsible for what the plane does. Wouldn't be substantially different for an automated car where control is still ultimately left up to the driver.
A wall has somehow acquired access to a computer and a keyboard....
If "JDS62F" corporation created an autonomous car and that product ran someone over, you can bet they would not say "oh well we can't sue a computer."
I think its time to bow out of this one, no new points are being brought up!
A wall has somehow acquired access to a computer and a keyboard....
If "JDS62F" corporation created an autonomous car and that product ran someone over, you can bet they would not say "oh well we can't sue a computer."
I think its time to bow out of this one, no new points are being brought up!
Clue time: even if nobody in the U.S. is willing to take a risk on the technology for liability concerns, the legal environment where the other 2 billion drivers on Earth live is a little different. The world doesn't stop outside your doorstep. Liability concerns have not significantly slowed the march of increasingly automated inputs on cars; rather, they've made driving safer on the whole.
You can rest assured the trial lawyers lobby will be heavy in favor of this. For very good reasons. Talk about low-hanging fruit, ripe for the picking...
Please register to post and access all features of our very popular forum. It is free and quick. Over $68,000 in prizes has already been given out to active posters on our forum. Additional giveaways are planned.
Detailed information about all U.S. cities, counties, and zip codes on our site: City-data.com.
Please register to participate in our discussions with 2 million other members - it's free and quick! Some forums can only be seen by registered members. After you create your account, you'll be able to customize options and access all our 15,000 new posts/day with fewer ads.
