Prius Autonomous Driving

I think it is unsupported aggrandizement.

Almost nothing to do with what car he chose to make 'self driving'.

 

are you serious jumbo?

that guys a genius, and these cars will be on the road en masse, sooner than later.

and i'll be glad when it happens. no more traffic jams, no more accidents, maiming, deaths.

no/low insurance

 

I cannot be impressed without a better understanding of the regulatory and safety oversight. Before this can be rolled out to consumers, the required levels of software verification and fault handling and crash prevention are quite high, but none of this is addressed in the article.

Put another way, don't bother showing me one successful cross-country trip. I need to see how the system can produce and verify a mortality rate of under 1 death per 100 million vehicles miles traveled, with a roadmap in place for pushing that rate down to less than 0.1 death/100MVT.

 

what is the current rate

 

1.25/100VMT.

But that is with the current mix including a lot of dangerous drivers and driver behaviors. I'm sure that the good drivers within the mix are considerably safer than that. Autonomous vehicles need to aim to at least match the good ones, hence the near-term 0.1 target. Plus a longer term target for 0.01.

On this scale, commercial airlines are now about 0.002. But they normally use a different scale based on departures or flight segments, not miles.

 

idk, if it saves one life, it's worth it

 

When spending billions of dollars, I want to see far more than one life saved. Otherwise, that money can save far more lives by spending it on other safety systems.

With a GDP of only $20 Trillion, and annual deaths running at 2.6 million, we simply don't have the resources and GDP to be spending a billion dollars to save a single life. In general, we shouldn't be spending much over $7 million per life saved, because doing so takes resources away from saving more lives elsewhere.

 

or takes away from worthless spending

 

Doesn't matter. Even funds taken from 'worthless' spending should be funneled to things that save more lives per dollar spent. And there is a lot of low lying fruit out there.

U.S. traffic deaths have exceeded 40,000 per year for several consecutive years, and serious injuries exceed 4 million. Anything that can take a significant bite out of that, say a 10% reduction, is worth many $billions.

 

true, but won't be. so if this guy can swing it, good luck to him. probably save more than one life

 

When these systems are actually ready, they should save many lives.

But they shouldn't be permitted for consumer sales as autonomous cars until they can at least match the average fatality rate of human drivers. And considering how many drivers are already abusing current driver-assist features, mentally disengaging with systems not sufficiently safe for such use, it is not at all obvious that the next round of 'autonomous' systems will be good enough to save any lives. If they promote too much driver disengagement, too soon, they could well cost many lives.

 

agreed. they need to be driver less.

 

The story about the test drive in Prime of first stage of new autonomous driving truck stack: Levandowski’s Pronto.ai plans to ship automated driving systems for trucks in 2019 – TechCrunch


iPhone ?

 

Banning all cars will save many lives. Is it worth it?

 

Proof of concept was done back during the DARPA Grand Challenge, more than a decade ago. The industry is beyond that now.

 

lets ban cell phones first, and see how it works out

 

Don't rush me...