Thank god the Aliens keep leaving us tech. Forget about the stones. This is far more useful, except for outages, solar events and non-wired disturbances which cause everyone in town to be injured simultaneously in dramatic fashion.
Well, there was no mention of whether this was achieved with or without the human occupant taking controls.
I find it hard to believe that the vehicles never encountered a situation such as the ones mentioned before, that required the passenger to intervene to prevent an accident from a jaywalker or another car doing something stupid.
The car on autopilot and getting bailed out to prevent an accidents isn't the same as the car operating completely autonomously and not getting into an accident. And if the occupant has to be as alert as if they were driving the vehicle (much like, say a c-train conductor), what's the point? It's cool, but if I can't be doing the crossword puzzle while I am being driven to work, it seems a bit silly. The only benefit would be the system monitoring and preventing the driver from making an error in judgement when changing lanes or possibly blowing through a red light.
Your contention then is that human reaction is quicker and more effective in a fast and dangerous situation than that of a computer? I'm not sure if you realize just how robust this technology is.
Yeah, an obvious exaggeration but if google tracks us on the web is it a stretch to think they'd also track our movements in their car, all for our own convenience of course?
Of course movement will be tracked, just like it is right now on your phone. If you go to google maps, you can see how long a commute along a particular route is at any given time because of phone signal tracking, or you can see this on Deerfoot with their signs that tell you how long it is taking to get to a particular exit. The negative impact on our lives will continue to be zero as nobody really cares where we are.
Your contention then is that human reaction is quicker and more effective in a fast and dangerous situation than that of a computer? I'm not sure if you realize just how robust this technology is.
Not quicker, but able to ascertain some risks. A system that prevents you from changing lanes when a car is in your blind spot is great, but I don't see how a fully autonomous system will be able to react (not only speed based) when 3 or 4 different things are going on. For example, a pedestrian jaywalks, some cars stop, some don't, you're either going to get rear-ended if you stop, and then the physics just of the environment and the ability of your car to react as expected.
I am saying that this is great to augment the driver experience, but the driver will still need to be fully alert. There will probably be no opporunity to sit back and take a nap or play Angry Birds while your car is driving. Like I said, even the C-train, that is on a track and limited in its route still needs a conductor.
Not quicker, but able to ascertain some risks. A system that prevents you from changing lanes when a car is in your blind spot is great, but I don't see how a fully autonomous system will be able to react (not only speed based) when 3 or 4 different things are going on. For example, a pedestrian jaywalks, some cars stop, some don't, you're either going to get rear-ended if you stop, and then the physics just of the environment and the ability of your car to react as expected.
I am saying that this is great to augment the driver experience, but the driver will still need to be fully alert. There will probably be no opporunity to sit back and take a nap or play Angry Birds while your car is driving. Like I said, even the C-train, that is on a track and limited in its route still needs a conductor.
I'm not sure what to tell you other than, you will soon?
Of course, anyone can come up with perfect storm situations in which accidents are unavoidable no matter what is driving your car. But the entire purpose of this technology is to analyze multiple things that are going on and react at speeds that would be untouchable by humans.
C trains are a weird comparison here as that don't run on the same tech, but yes, within 10-12 years, you will be a complete passenger in a car driving itself, being able to read, play video games, do work, or whatever. You will look out the window and see cars driving with no one in them, and you will be at ease knowing that this is the safest you've ever been inside of a car by a considerable margin.
The Following User Says Thank You to Major Major For This Useful Post:
Well, there was no mention of whether this was achieved with or without the human occupant taking controls.
I find it hard to believe that the vehicles never encountered a situation such as the ones mentioned before, that required the passenger to intervene to prevent an accident from a jaywalker or another car doing something stupid.
The car on autopilot and getting bailed out to prevent an accidents isn't the same as the car operating completely autonomously and not getting into an accident. And if the occupant has to be as alert as if they were driving the vehicle (much like, say a c-train conductor), what's the point? It's cool, but if I can't be doing the crossword puzzle while I am being driven to work, it seems a bit silly. The only benefit would be the system monitoring and preventing the driver from making an error in judgement when changing lanes or possibly blowing through a red light.
The test vehicles did have times where a human did have to intervene. That's what makes it a test and how it learns. The newest prototype doesn't even have a steering wheel or brakes, so it's a moot point in the future, but I see your concern now. The video I posted earlier gives a good run down on the technology.
Any way you look at it, there will be less accidents, injuries and deaths than there is now. There has not been a single new health intervention that would have such a health impact since maybe vaccination.
Think about this: There were 9.5M motor vehicle accidents in the US in 2009. Because of those accidents, 2.2M people were injured. That's an awful lot of injuries to prevent even if there is the odd accident
Yeah, it's not a question of can a computer react appropriately in all situations, it's a question of can a computer react better than most people in all situations. That's a pretty low bar lol.
__________________ Uncertainty is an uncomfortable position.
But certainty is an absurd one.
The Following 8 Users Say Thank You to photon For This Useful Post:
I can't wait until the day when no one owns their own car. There is no need. For most people they only are actually using their car at most 2 hours of every day. The rest of the time it sits in your garage or a parking lot.
It will be like a taxi that picks you up almost instantly and drops you off at the door of where you need to go. Need to keep your things in it for a time period then let it know that and it won't go pick someone else up and wait nearby for you.
Even if they are self-driving, I think people are still going to want to own their own cars. Even if (or especially because?) you have nothing "to do", a lot of people are still going to want some choices, personality, luxury, and something to show off.
Even if they are self-driving, I think people are still going to want to own their own cars. Even if (or especially because?) you have nothing "to do", a lot of people are still going to want some choices, personality, luxury, and something to show off.
I think you're right, but it may come down to being a luxury rather than a necessity
Location: Close enough to make a beer run during a TV timeout
Exp:
I'd be curious to see how it does with judgement calls. A few situations I have been in where I had to choose the lesser evil:
- Going down the freeway in Florida, and traffic goes from 75 to 0 immediately. All cars jam on their brakes, and I hesitate as I see the 5 ton truck behind me is unable to stop. Would the Google car just stop to avoid the car in front, or would it consider the truck behind me as well.
- Similar situation on Deerfoot. I veered into the breakdown lane and the guy behind me plowed into the guy who was in front of me. Once again, would the Google car just stop as quick as possible.
- After a multi-car collision, I was heading off a cliff. I managed to get the car to fall off at a point where the drop was only 20 feet, and either side was at least 200. Could it factor for that?
- One thing that didn't happen to me, but given the choice after losing control, I might be inclined to hit a larger vehicle instead of hitting pedestrians. How would the Google car react?
I really think it's cool, and if I could drive overnight on a road trip and arrive at my destination rested, and with my vehicle, I would be signing up. I'd just like to know how it would react to odd situations.
The Following 3 Users Say Thank You to ken0042 For This Useful Post:
Even if they are self-driving, I think people are still going to want to own their own cars. Even if (or especially because?) you have nothing "to do", a lot of people are still going to want some choices, personality, luxury, and something to show off.
But to that point, I'd say why do people (assuming all other things are equal such as time and cost) prefer take a train where they need not worry about driving and just be whisked to their destination.
What I think will be the determining factor is whether or not the car can learn to react in adverse weather conditions. Considering the car relies on "seeing" the roadway, the car would fail in Canada about 3/4 of the year due to all the snow. As well, how does insurance work? If you slam into another car, who is at fault? The passenger? The owner of the car? The manufacturer?
But ultimately, my cafe-sippin (yop-gobble avoiding) brain is worried that a driverless car is just a progress trap, enabling people to continue to live individualistic lives and continue isolating people from everything else around them. However, should the technology be used wisely, it may have great merit.
My one open-ended question is why do we need to use roadways? Instead of paving roadways that constantly require maintenance due to potholes and other wear and tear, why not look at innovating alternatives? Maybe a monorail-like network of personal pods milling autonomously about the city (and region (and country)) would be better? It would use less ground space, arguably cost less on a per unit basis (save the fact you'd have to start from scratch), immediately solve the pothole problem, and have far fewer variables at play.
Yeah I've had a few similar situations where I had to swerve out of the lane to prevent getting rear ended. Stuff like that should be straight forward to model and consider as options, even modelling how long larger vehicles take to stop.
The cliff one is interesting, you should almost email that to them and see what they say, maybe it's something that's possible but they haven't thought of yet.
Other stuff too like road conditions. Can it detect wet pavement vs dry, or is there loose gravel scattered on the pavement.
__________________ Uncertainty is an uncomfortable position.
But certainty is an absurd one.
Yeah I've had a few similar situations where I had to swerve out of the lane to prevent getting rear ended. Stuff like that should be straight forward to model and consider as options, even modelling how long larger vehicles take to stop.
The cliff one is interesting, you should almost email that to them and see what they say, maybe it's something that's possible but they haven't thought of yet.
Other stuff too like road conditions. Can it detect wet pavement vs dry, or is there loose gravel scattered on the pavement.
I think this is a pretty interesting example of the 'human' element.
What matters isn't what the materials or conditions are, but how the vehicle is reacting as a result of them.
Between electric vehicles and advanced computer systems already in used in a variety of vehicles, the road conditions are secondary to the output of the vehicle.
For examples, whether it's gravel or wet does not really matter, the car is adjusting the power and direction of the vehicle in relation to it's place on the road. The car may not 'sense' water, but it will sense traction reduction or uneven power distribution and react accordingly. The same will be true for breaking, steering and essentially every other component of the vehicle.
This is a rudimentary introduction to how prevalent much of this technology is in our vehicles already.
I love this clip.
The Following User Says Thank You to Flash Walken For This Useful Post:
But to that point, I'd say why do people (assuming all other things are equal such as time and cost) prefer take a train where they need not worry about driving and just be whisked to their destination.
What I think will be the determining factor is whether or not the car can learn to react in adverse weather conditions. Considering the car relies on "seeing" the roadway, the car would fail in Canada about 3/4 of the year due to all the snow. As well, how does insurance work? If you slam into another car, who is at fault? The passenger? The owner of the car? The manufacturer?
But ultimately, my cafe-sippin (yop-gobble avoiding) brain is worried that a driverless car is just a progress trap, enabling people to continue to live individualistic lives and continue isolating people from everything else around them. However, should the technology be used wisely, it may have great merit.
My one open-ended question is why do we need to use roadways? Instead of paving roadways that constantly require maintenance due to potholes and other wear and tear, why not look at innovating alternatives? Maybe a monorail-like network of personal pods milling autonomously about the city (and region (and country)) would be better? It would use less ground space, arguably cost less on a per unit basis (save the fact you'd have to start from scratch), immediately solve the pothole problem, and have far fewer variables at play.
Vehicles have the potential to view their surroundings in superior way than we do, through a variety of visual spectrum. Utilizing 'vision' beyond the human visual spectrum would in and of itself reduce many of the dangers you're talking about like heavy rain, snow and fog.
The thing is, we already have roadways. We might as well use them. Computer controlled vehicles would go a long way towards reducing maintenance costs on those roads as well.
The other thing that driver-less cars do is reduce many of the variables humans see as obstacles for the driver-less car to overcome. Standardizing road travel would go such a long way in reducing public expenditures of roadways, private expenditures on maintenance and fuel for vehicles and have an incredibly beneficial effect on public transportation.
There was an episode of Top Gear where one of the hosts had to drive an exceptional distance on a tank of gas, further than the listed range of the vehicle by quite a bit. To do it, he had to pay constant attention to the road to maximized his fuel efficiency, including avoidance of using the brake and gas, giving appropriate room on the road for other users and generally hyper focusing on his journey.
He described it as the most mentally exhausting thing he'd ever done, but he managed to double the fuel economy of the car, and of course, no accidents.
Imagine if that was so effortless you could surf calgarypuck while it was happening?
Even if we only look at the logistics of an autonomous car driving in ideal weather conditions, I believe the system only really comes together well when you then remove all human elements. To have one autonomous car in an ocean of many human piloted cars would be a disaster. This works better if all, or most cars are autonomous and aware of each other's route, position, destination, etc. That's when it will be able to shine. When you eliminate the drivers that pull a u-turn or some non - standard driving maneuvers from the equation then this would be really magnificent. As Devil's advocate, it seems that step is a long way out.
The Following User Says Thank You to Wormius For This Useful Post:
I can see the advantages in having a driver less car in some situations but the thing is I enjoy driving. It seems like many of you don't.
Me too. I love driving, and it seems less and less people agree as the years go by. Maybe it has to do with getting older. I'm 29 and still enjoy the feeling of being in control of a vehicle. Maybe that will change as I get into my 30's.
