So Alex Roy started a group to protect against a world with only autonomous driving… in case we ever get there.
Roy started the Human Driving Association in early 2018, after reading a manifesto written by Robin Chase, the entrepreneur and former C.E.O. of Zipcar. In the manifesto, “Shared Mobility Principles for Livable Cities,” Chase outlines a now familiar vision for autonomous driving, in which dense urban areas are filled with autonomous vehicles that operate in fleets, so as to reduce congestion and emissions. “I just couldn’t understand why it had to be so binary—individual ownership or shared fleets,” Roy said. In ninety frenzied minutes, he wrote his own manifesto, which he published as an article for The Drive, an online car magazine. It begins with a picture of a steering wheel and the words “From my cold, dead hands,” and culminates by calling for a constitutional amendment creating a right to drive.
Interesting read but a law won’t save the human piloted car. The move to autonomous will not come from a law banning drivers, since it would never pass, but from the insurance industry. As more and more autonomous cars hit the roads and continue to prove their safety superiority to human drivers insurance companies will just keep raising the rates on non-autonomous cars until they become luxury toys of the rich.
I’m against creating “rights” like this that are not inherit in nature.
But I think autonomous driving will really impact cities first, so maybe driving won’t just be “luxury toys for the rich” but also for those who live is sparcly populated areas.
It will start there for sure, but crash data will push insurance companies to penalize human drivers big time if the computer driven car proves to be as good a driver as the early data shows. Once that starts happening it will be a pretty fast changeover to AI cars. You have to remember that a large chunk of “the masses” don’t care about cars. They just want to get from A to B without taking a stinky bus/train with the rest of “the masses”. They’ll quickly abandon their steering wheels when their yearly insurance for their car goes up 50% while the insurance on an AI car drops. Once you lose them manufacturers lose the incentive to build steering wheel cars in the first place so the R&D shifts to AI vehicles. This change will happen fast once it starts. Anyone remember manual transmissions? Same thing, but this time you’ll have insurance pushing it too.
I think @JayS is right. The transfer to AI only driven vehicles will be driven by money.
I wonder how speedtrap towns :cough: Corfu will be able to afford to stay incorporated once cars no longer speed…
Shredd and Reagan were talking about how they want to add anti speeding features to cars. The gps in the car won’t allow you to speed on the road you are on.
Anyone know how Google’s autonomous system does in a snow storm?
Thinking about the sensors and programming required to have an AI car know the intersection coming up looks icy the way a human can makes me think full AI nationwide is a LONG way off still.
I would think that forward looking infrared would be able to see that even better than a human. I think Mercedes is already using this with its HUD to detect animals ahead. Add to that the internet of things, the ability of the cars to talk to each other, other cars in the area can alert each other of adverse road conditions. IE Once one car hits a pothole all the others will attempt to miss it…
Didn’t even consider this, especially if it’s as accident free as they say it’s gonna be.
EDIT- @LZ1 besides the collision industry think of how many less cars they’re going to sell, even if the manufacturers are the ones making the autonomous ones.
Yea the only accidents that will happen will be when the NS-5 robots jump out of their automated semi-trucks and try to kill the people who realize their plans.
Edit @ubengineering I also think sales will eventually hit a plateau because of diminishing returns and product longevity. With composites and aluminum becoming prevalent corrosion becomes a non issue. Electric motors have theoretical life times of over a million miles. And once the car drives itself and has nearly every comfort feature what incentive is there to buy a new car? I’m reminded of how I used to get a new cellphone every year, now the differences are so small its every three years.
Yeah, I’m not thinking so much as the ability to see through falling snow to see obstacles like cars as I am the ability to see the lanes, shoulder etc. From what I’ve seen of the testing of these AI systems they rely heavily on reading the lines on the road.
The other big programming obstacle is being able to handle the quickly changing road conditions and adjusting braking zones and cornering speeds appropriately. Picture the crappy plowing around here where you turn off a plowed down to pavement road into a side street with 6" of untouched snow or slushy ruts Then picture the idiots that blow snow in the street when clearing their driveways where you’ve got clear dry pavement and all of a sudden 3" of crap. Or worse yet, black ice. That braking zone that only needed 100’ of moderate braking to stop before hitting the car at the red light suddenly needs 200’ ABS fully engaged braking to stop. Even experienced drivers with all the finely tuned senses a human has have trouble seeing that black ice and judging an appropriate stopping distance.
Car & Driver had a nice article on this that I actually just happened to be reading the other week. Some pretty fascinating stuff and touches on a lot of the points already mentioned. It will really be interesting to see how the faults play out in an insurance situation. That along with “perfecting” the autonomous driving experience has me thinking the concept of fully autonomous roadways are still a long long ways away.
My method of “testing the waters” is constant ABS engagement and adjusting speed. But this is usually when my speed is a “little” north of posted.
For programming I’d like to see V2V data within the last X minutes and single wheel slip monitoring by two methods:
-Single wheel braking, to measure mμ, this would be slightly intrusive, but less so than my 4 wheel ABS method. Ideally a 4 motor car would be great at constant monitoring.
-5th wheel, a simplified / smaller version of a Saab airport tester.
