The Passionate Eye: Driverless Cars

[khutchi]

Very interesting CBC episode on the immediacy of high level driverless autonoumous vehicles and the ramifications for our cities. The removal of traffic lights, Less need for public parking, charge sharing networks, less emergency room health care demands. All by 2021?

http://www.cbc.ca/passionateeye/episode ... erless-car

Only available for a limited time. Check it out!

[Dizzy1]

Not going to happen, not for a long time still. The technology simply isn't even being close to safe or accurate enough to unleash completely on the public. The legal aspect (liability, etc) is the real hurdle though.

Anyone who has a car that semi-autonomous, things like self-parking, lane assist and adaptive cruise control, know that there is still a lot of work to go before they're reliable enough to achieve a driverless utopia.

And personally, as a person who loves to drive, hopefully not in my lifetime because that will be a very, very bleak world indeed.

[Dizzy1]

Another very interesting point of view ...

https://www.cnbc.com/2017/02/27/buffett ... ustry.html

[khutchi]

Another very interesting point of view ...

https://www.cnbc.com/2017/02/27/buffett ... ustry.html

Always wondered what would happen to my $200/month ICBC premium when my car doesn't even have a steering wheel anymore

[GordonH]

In the possible rare chase of an auto accident involving autonoumous vehicles, who would get charged..... manufacturer. Can't charge the owner since he/she had no control.
Lawyers are going to love it, until all bugs are dealt with.

[Veovis]

In the possible rare chase of an auto accident involving autonoumous vehicles, who would get charged..... manufacturer. Can't charge the owner since he/she had no control.
Lawyers are going to love it, until all bugs are dealt with.

Give police an override......of course that gets hacked in a week and people are car jacked and redirected to robbery spots with no control.

I love the increase of tech, but realities still exist that people like to ignore for the "sounds cool" factor.....few CEO's do though.

[Queen K]

Would one trust a driverless car explicitly? I'm not a perfect driver, but I can guage actions around me that technology can not.

Not sold on the idea, but I can see where the aging population and the boomers are creating a demand.

[Dizzy1]

Would one trust a driverless car explicitly? I'm not a perfect driver, but I can guage actions around me that technology can not.

Take adaptive cruise control as an example - you're driving down a multilane highway with the cruise set, you're in the left lane about to pass a car in the right lane, the road bends to the left, the car detects an object in front of you (all though its not really in front) and puts the brakes on.

Lane assist - great feature, but it doesn't always work - when the road markings aren't detectable by the car's camera, it doesn't work.

Sensors/cameras - all these systems rely 100% on sensors and cameras built into the car - if one sensor fails or gets covered - thats it, the system no longer works.

Speed limits - the car will have to have a GPS installed, and all speed zones will have to be updated to reflect current speed limits . Many nav systems show you the speed limit on the road you're driving on, but they're not always accurate (updated) and they don't show construction zones. So what happens if there is a construction zone set up overnight? They'd have to install transmitters that automatically update every car on the road in real time.

Its a fun idea but its a lot more complex than most people realize.

[StraitTalk]

Would one trust a driverless car explicitly? I'm not a perfect driver, but I can guage actions around me that technology can not.

Take adaptive cruise control as an example - you're driving down a multilane highway with the cruise set, you're in the left lane about to pass a car in the right lane, the road bends to the left, the car detects an object in front of you (all though its not really in front) and puts the brakes on.

Lane assist - great feature, but it doesn't always work - when the road markings aren't detectable by the car's camera, it doesn't work.

Sensors/cameras - all these systems rely 100% on sensors and cameras built into the car - if one sensor fails or gets covered - thats it, the system no longer works.

Speed limits - the car will have to have a GPS installed, and all speed zones will have to be updated to reflect current speed limits . Many nav systems show you the speed limit on the road you're driving on, but they're not always accurate (updated) and they don't show construction zones. So what happens if there is a construction zone set up overnight? They'd have to install transmitters that automatically update every car on the road in real time.

Its a fun idea but its a lot more complex than most people realize.

The companies that are making progress in this field are not relying on updates and databases of information so much as they are creating AI which acts and reacts in real-time, with eyes and sensors. What I've seen is pretty astounding. Autonomous vehicles aren't necessarily good for the automotive industry though, so I don't see this happening soon unless the Tesla's of the world start churning out really cheap cars.

[Dizzy1]

The companies that are making progress in this field are not relying on updates and databases of information so much as they are creating AI which acts and reacts in real-time, with eyes and sensors. What I've seen is pretty astounding. Autonomous vehicles aren't necessarily good for the automotive industry though, so I don't see this happening soon unless the Tesla's of the world start churning out really cheap cars.

The problem with cameras and sensors is that it can only recognize what its been programmed to recognize and react in a way its been programmed to. If there is no standard in speed limit or other regulatory signs in regards to placement, shape and design, it won't be nearly as accurate or reliable as a GPS based system would be. Add to the problem, if a sign is obscured by dirt or other debris and becomes unreadable, or if the cameras and sensors get covered in dirt, the system won't work. Also, every manufacturer will have their own platform, programmed differently with no set standards.

Don't get me wrong, we're getting there, but its still a ways off. Its not really a question of the vehicle technology, its standardizing the infrastructure that will be the real hurdle.

Looking at the aviation industry - aircraft are fully capable of flying themselves, and have been for quite some time, but the industry and regulators still understand the importance of the human interaction involved when things go south. Human's, as flawed as we are, are still able to react better in abnormal situations then automated systems. Having said that though, a lot of that has to do with the training we receive in operation of a vehicle, which of course, when it comes to cars we are seriously lacking.