http://www.cbc.ca/news/canada/kitchener-waterloo/ontario-announcement-automated-cars-1.3268181
Testing to start as early as the New Year.
Testing to start as early as the New Year.
Would you feel safe sharing the road with a robot car?
- Thread starter rafterman
- Start date
Not today. But give this technology 20 years to mature then definitely. Looking forward to never having to worry about drunk drivers on the road.
Holy smokes, I fucking certainly don't feel safe with meatballs shaving on the 401 right now. bring on the robots for sure
Yes. Unless there's some sort of malfunction I would feel safer having robot cars on the road than the current morons on our highways.
I guess it depends on why it got hit. How well does it see pedestrians or cyclists or school buses?
I imagine they'd get passed a lot, if they are programmed to drive with a safe distance between cars they'd have been butting in all the time, forcing them to drive slower.
But then again, if I could relax and not worry about the car on longer drives I would be ok with it taking a little longer.
We're a long way from reliable, affordable self driving cars. I could be wrong but I still say there's way too many variables to make it safe and reliable. These things use a variety of sensors including laser and radar. How do sensors perform in snow and rain? What if it's raining heavily and the vehicle approaches an intersection but the traffic lights are out. There's cops on point duty, can it follow their instructions? What if there's a detour? Can the vehicle read the signs and follow them accordingly? What about sign laws, no left turn 4-6 PM M-F? Can it read the signs and act accordingly?
Here's some of the problems they have to iron out. At the moment they're huge obstacles.
Here's some of the problems they have to iron out. At the moment they're huge obstacles.
@GameBoy27 is right. Lots of issues with full autonomous vehicles. I think complete, reliable door to door self driving cars is at least 15 years away.
In the meantime, cars are going to come out with tons of autonomous features that will still require a driver to pay attention and take over if necessary. Self driving on the highway, self driving bumper to bumper traffic on the Gardiner. Self parking.
Should be cool.
In the meantime, cars are going to come out with tons of autonomous features that will still require a driver to pay attention and take over if necessary. Self driving on the highway, self driving bumper to bumper traffic on the Gardiner. Self parking.
Should be cool.
I can (mostly) predict what the idiots on the road are going to do, and respond before they do it.
Robots, I don't know. I worry that in their quest to do the "safe" and "right" thing they will do something that surprises me and causes unintended consequences.
Robots, I don't know. I worry that in their quest to do the "safe" and "right" thing they will do something that surprises me and causes unintended consequences.
Not are going to, have. Lots of 2015/2016 models have numerous self driving capabilities billed as driver assistance or safety features: blind spot detection, forward collision avoidance, adaptive cruise control.
Many cars now have cameras that watch the car in front and adjust speed to maintain distance, while also adjusting the steering to keep you in the lane. At that point you as driver could fall asleep and the car would just keep driving, except some models also watch your face and yank your seatbelt if they detect you are inattentive.
They don't do lane changes, tur but they will scream at you if you attempt an unsafe lane change, etc.
The technology is really close.
All that stuff you mentioned makes driving safer, which is a good thing. I'm all for technology, but completely driverless is so far away it's not even funny.
Did you read how far away they still are? http://www.slate.com/articles/technology/technology/2014/10/google_self_driving_car_it_may_never_actually_happen.html
Making it work everywhere, practical and affordable is decades away. Besides, it will likely only serve a niche market.
Initially it will serve only a niche market, like initially mobile phones served only a niche market, and computers served only a niche market. But as their capabilities grow they will eventually spread out and dominate everything. Cars with steering wheels will eventually be the niche market.
I expect the first niche market to be taxi service in urban cores. The streets could be well mapped, by some ultra version of google streetview that maps every inch, maybe even have special traffic controls installed, and the speeds kept low (no highways) and there would be a lot of demand--you call uber, this self driving car comes to get you, and takes you to where you're going while you read a book. In Toronto imagine it covering the city south of Sheppard between Dufferin and the DVP. You'd need a driver to go to the burbs but there would be a lot of people moving around in that zone, from home to club, office to client, home to work, etc., and there's a substantial savings in eliminating the driver--once the technology costs less than an annual salary, it'll be profitable.
edit --
Read your article, and that is why I think the first niche market will be the urban taxi service. They CAN map, and they can co-ordinate with the city to install self-driving-friendly traffic signals, avoid school areas, etc., and still deliver a useful service. The claim that they are a long way off because you would have to map every square inch of every US road is wrong--that won't be the first use case. The first use case will be self-driving cars on specific prepared roads where the municipal government coordinates with the mapping service when they put up a new traffic light.
The self driving taxis could also follow something like a bus route, except you can customize it--certain mapped out, properly signaled, prepared streets that it sticks to and drops you as close as it can to your destination.
Highways will be the next use case for the same reason--a 400 series highway can be fully mapped, end to end. You drive to the highway, take a nap for the bulk of your inter-city trip, and wake up to drive the car around once you get off the highway. Unless you get off in downtown Toronto, in which case, the car may be able to handle that. Something like the US interstate system could easily be fully mapped long before your sidestreet is.
But yeah the idea that they'd be able to drive on unmapped, unprepared roads is a long way off--but then, once upon a time, there were only a few roads, most of them unpaved, and only a few cross country routes, and people still found lots of uses for cars (e.g, the famous "route 66"). The self-driving car will be like that at first--only go on certain roads, human takes over off that, and it will be very useful. Car that allows you to fall asleep at the wheel on a long trip. Cheaper taxis. These will happen before they just drive everywhere.
We all know eventually this is something possible
In today's world it is an accident waiting to happen
Robo-car only works if road is in good repair & properly marked
The shit show that is GTA roads constantly under construction & mispainted lines.
My car right now has a fit because the many construction zones have mispainted yellow/white lines, I would NEVER give a car total control just based on what I have seen the drift avoidance software does.
The idea is to stay inside the lines, well what about when lines go into a wall?
In today's world it is an accident waiting to happen
Robo-car only works if road is in good repair & properly marked
The shit show that is GTA roads constantly under construction & mispainted lines.
My car right now has a fit because the many construction zones have mispainted yellow/white lines, I would NEVER give a car total control just based on what I have seen the drift avoidance software does.
The idea is to stay inside the lines, well what about when lines go into a wall?
Check out this video of a Tesla Model S with the latest software update being released this week. The video shows the car driving itself and the driver changing lanes just by flicking the turn signal.
https://www.youtube.com/watch?v=Cw7Esg-txR0&feature=youtu.be
https://www.youtube.com/watch?v=Cw7Esg-txR0&feature=youtu.be
Or what about when it's snowing and all the lines disappear?
Who in their right mind would buy a Tesla Model S only to have it drive you around at exactly the speed limit, never accelerating or braking beyond the minimum requirement. Talk about a waste of a beautifully engineered car that's meant to be "driven".
I think its further away than they are saying. Construction, signals out, pedestrians, cyclists, passing streetcars, snow etc- I don't believe they can deal with all this and hackers.
But is it economical? Why would uber buy cars when there model is to be a back office service provider ( a taxi pimp if you will) rather than a capital intensive business of buying and maintaining a fleet of cars? Its contrary to their whole business model
I think their new autopilot software is a perfect blend. It's only there when you want it. Boring highway commutes and traffic. It's not limited to the speed limit, you can set it to go any speed you want.
