Autonomous cars, unevenly distributed


#1

[Permalink]


#2

I would really be curious to see the results of rephrasing the question to " would you rather buy an autonomous car from the largest advertising company in the world (Google) or a company that has been building conventional cars for over 100 years (GM)?"


#3

Is there some compelling social benefit (of the kind used to justify special treatment for HOV or electric vehicles) that would justify ‘special lanes’ for autonomous vehicles, especially while they are still in the ‘pricey option’ stage?

Sure would help uptake; but subsidies tend to do that…


#4

I wonder what the market is in China for autonomous vehicles?


#5

Autonomous cars drive more predictably than humans, have faster reaction times, don’t get sleepy or drunk or hungry or distracted by texting, and don’t fall prey to stupid human psychology (they’ll happily drive steadily at the speed the road can bear, where humans will speed up as soon as there’s a clear stretch only to brake at the next traffic jam caused by people driving too fast for the road).

A lane in which all the cars could verify that all their neighbours are computer-driven could allow cars to safely with a closer following distance, reduce congestion and stop-and-go traffic jams, etc., by driving in ways that humans either can’t or won’t.


#6

That would negate one of the benefits of auto-driving cars though. Simulations have shown that replacing 25% of the cars on the road with auto-driving, or even adaptive cruise control cars eliminates traffic jams, since there are enough automated cars in the mix to make the other 75% drive less stupider.

Edit - My memory is bad. According to this report at U Mich it’s only 20% that’s required.


#7

I hadn’t thought of that. That’s a good point - if you’re stuck behind a
sufficient contingent of cars driven at the optimal speed for the current
road conditions, you’re pretty much forced to drive sensibly yourself,
unless you’re willing to do some serious idiot manoeuvres.


#8

The thing that I always wonder about is merging. So few people know how to do it properly manually that I assume the human drivers would mess up the automated cars algorithms something awful.


#9

There are a lot of maneuvers that are difficult to do, but wouldn’t be if everyone was driving predictably and rationally. Merging definitely falls into that category. I suspect that human drivers will find themselves with bigger margins of error as the autocars give them a wide berth.

I am more curious about road conditions. What about black ice? Slippery mud? Sleet? High winds? Icy roads in the mountains? The slushy/icy/snowy roads we often get in residential areas that aren’t plowed often in winter? Presumably the cars would be able to handle most of that - much better than the majority of humans in my opinion. But in order to be insurable they will have to be almost foolproof at it, which is another level altogether.

I can already see the next generation of Hollywood car chases involving human driven vehicles blowing through crowds of autocars which frantically get out of the way.


#10

Except Google is more of an Artificial Intelligence / Machine Learning company that makes money by applying AI/ML to advertising.


#11

It seems more likely that you would be buying a car, made by GM, Now with Google driving technology!!! Buying Motorola was kind of a stretch for them, they aren’t going to start making anything more than an on-board computer for cars.


#12

True, I never actually considered the possibility that Google would start manufacturing cars. I just assumed they are doing their programming and testing on cars that are already in production. I can’t imagine it being even remotely profitable for them to get into manufacturing at any scale really.

The money for them will be in selling the software and technology, and being first movers. Ford, Toyota and all the rest are car makers - they will just implement the stuff. I don’t see Google producing cars any more than I see Hyundai getting into the online search engine market.


#13

Cars are already at an advantage there; traction control has been implemented in ordinary cars for ages now, and it gets more sophisticated every year. That has sort of bugged me for a long time, in that it has coddled bad drivers and not given them any real incentive to improve their skill set.

All this has been considered. Automated cars can simply drive better than we can, and all that remains is for us car-buyers (and legislators) to get used to the idea.


#14

I’m with you there. I’ve never had an accident (after 27 years of driving). Part of that is attributable to me as a careful driver (my parents owned a driver training school and we had an infuriating passenger side brake in the car I learned with). At least some of that is purest luck - there have been near misses when I wasn’t paying attention, didn’t see the other car or whatever. I am a good driver, but I have little doubt a computer has a better attention span than I do.

I think the major entry point for the technology will be in transport trucking. Because removing the driver altogether, at least on main routes, will be profitable. Taxis won’t be far behind (or could lead the way). They will require payment before leaving or something, but again will take the biggest cost out of the equation.


#15

This topic was automatically closed after 5 days. New replies are no longer allowed.