base2 

Then you haven't been keeping up on the topic. They've been field tested for years. On open roads in active communities, dense urban environments, & open highways, nationwide. So far zero "accident" has found the car to be in violation of any traffic laws. And their current average time between collisions is roughly 1/10 the average drivers.

Progress has been truly staggering.

OBoile

You should have just stopped here...

I-Like-To-Bike

The hype is indeed staggering; effective too!

Source or reference for data about collsion information for autonomous cars when operated without human oversight on open roads in active communities, dense urban environments & open highways, nationwide? Maybe even a picture of one of these "autonomous cars" that has been "proven safe" to be operated without human oversight on open roads in active communities, dense urban environments, & open highways, nationwide.

JW Fas

Or, insurance companies will charge nominal pricing for AVs and make manually controlled vehicles prohibitively expensive to insure. That seems to be the more likely scenario given insurance companies exist to profit as much as possible, and pricing will reflect what the market will bear.

rossiny

How do you know that if AV, do in fact come into production, that roads will not be modified for their use and in turn ban cyclists from those roads?

Daniel4

You stepped right into it, didn't you? I already knew what the response was going to be regardless of your references.

Don't feed the trolls.

I-Like-To-Bike

Right, regardless of the facts, or lack of same, just keep on playing a broken record, maybe it will sound true if repeated often enough.

I-Like-To-Bike

Why is this alleged factoid inevitable? Because you say so? Or because the hucksters and hypsters promoting their various schemes say so?

Rollfast

I've read articles lately that pretty much say that the Big Guys like GM are just abandoning the whole thing.

No money coming in. I think you will see it more in movies that on the street where you live.

livedarklions

My understanding is that the testing is going quite badly for reasons that aren't related to the fatal accidents. In real world testing, the AVs tend to freeze when they can't apply the traffic rules. For example, they tend to just sit in the intersection when they need to make a left turn in traffic. There's enough of these situations where our real way of performing the maneuver is really just figuring out what the other guy is going to do, that these vehicles are, so far, not ready for prime time as they would just clog the roads as indecisive drivers blocking lanes.

They could, at least theoretically, fix the safety issues, but still end up with a useless technology. After all, the most effective way to prevent collision is for the vehicle to never move.

Leisesturm

I am going to pay you a rare compliment ILTB and observe that you clearly possess an above average intellect. Therefore it baffles me completely that you have decided to come down hard on the contrarion POV on the whole AV issue. It is looney tunes to imagine that AV's aren't already much better drivers than 8/10 of the human population. Skillswise that is. Machine reactions are faster, period, no contest. The sensor suite of the advanced systems can 'see' in wavelengths of energy impossible for human senses to duplicate. The one thing they lack is intelligence. Real world driving demands tons of it. It's a low bar for us but remains a thus far insurmountable one for the AV. But to bet against them in the naive way that you do everytime the issue comes up is to ignore the evidence of the PC. The Internet. GPS. Space flight. Hell, commercial aviation. These are all commonplace technologies some of which did not even exist as concepts before the 20th Century.

Leisesturm

What you are saying in so many words is that AV and non-AV cannot presently co-exist in the real world. Actually, my understanding is that they are doing much better than your bleak assessment. But it is my own bleak assessment that IF an outcome that completely satisfies the DOT cannot be reached it is the deeper pockets of tech innovators and their Venture Capitalists that will prevail. Non-automated vehicles will be severely restricted as to when and where they can operate so as not to impact the safety of automated carriers. This is one Genie that is not going to go quietly back into the bottle.

Leisesturm

Abandoning what, exactly? Carmaking? Surely not! Automated carmaking? I don't know that they ever were much invested in that niche. But don't think that that isn't changing. GM or Ford or somebody has just caused a lot of pain to thousands of mid and high level employees as they restructure to focus on exactly the new opportunities in AV making.

I can't tell you what you've read but what I have read is that the Ford F150 pickup is still the best selling vehicle in the world. Period. But it is a fact that America's Middle Class is on life-support. The future buyers of F150's and a lot of things besides will not be American. After sucking us dry the Big 3 automakers are focusing their efforts (and factories) on India and China where the new Middle Class by the numbers anyway will simply eclipse anything America ever had or will have. But ... at the end of the day the V in AV is for Vehicle. Motor Vehicle to be exact. Yes, tech startups seem to be driving the AV industry but it seems to me like the companies best placed to provide the chassis and drivetrain platforms that the tech companies can overlay with sensor suites and guidance systems are the Big 3 themselves.

I simply can't believe that they aren't secretly up to their eyeballs in the design and manufacture of rolling stock for Waymo and Uber et al. My only regret is that no one seems to be taking advantage of this period of transition to also migrate away from the internal combustion paradigm. That the AV under testing as we 'speak' are not at least hybrid if not fully plug-in electric is crazy. Short sighted and crazy. We should be very upset about another generation of internal combustion engines destroying what is left of our environment that our children and grandchildren are going to have to live in.

tyrion

Because insurance companies charge less when risk is reduced.

base2 

You see, the thing is though is all the information you seek is available & public as a matter of law.

You are free to remain in the fog of your inability to Google if you choose. It is your obligation to present the evidence since you are the one with the outlandish claim.

But here, let me help. Evidence, in part of my claim: All California Autonomous vehicle accident reports.

Here, check this link: Cool stuff.

All by it self

Here is a link from 18 months ago. Maybe you define San Francisco's city streets as strictly controlled & regulated testing environment? What about Phoenix?...Asking for a friend.

Autonomous vehicles are here, normal, accepted, and progressing at tremendous rate in the real world. You just are unaware because of how seamless and relatively mishap free the transition has been. Welcome to 2019. It's awesome here.

I-Like-To-Bike

Sorry, but you are the fellow who claims that Autonomous vehicles are here, normal, accepted, and progressing at tremendous rate in the real world. None of your so-called "evidence" refers to motor vehicles being driven on roads without human oversight, ya know real autonomous cars, nor are press releases from the promoters evidence about vehicles "proven" to be capable of safely transporting passengers in dense urban environments & open highways in all weather conditions, nationwide without a human overseeing the operation at all times.

Your claim that "autonomous vehicles have proven to be much much safer" [than cars with human drivers] and "their current average time between collisions is roughly 1/10 the average drivers" is fabricated nonsense since so-called "autonomous cars" are being tested only in limited tightly controlled areas, weather conditions and with human oversight.

Your reference to California accident data is ridiculous as it is a comparison of extremely limited data on incidents involving test vehicles with human oversight (apples) with who knows what data on real world vehicles driven in who knows what conditions or environment (oranges.)

nayr497

Our local, elite university just shot down a lightrail project that was 15 years in the making. Their reasons were hard to decipher. Great timing though, as our once sleepy city is now booming and traffic has gotten extremely bad in a really short time. Not lot we need anything to mitigate that...

I-Like-To-Bike

True and the determination of risk and associated/appropriate rates are based on real world experience and data, not on the hyperbole and unsubstantiated promises of promoters, speculators and fan boys.

tyrion

Correct.

base2 

You asked for evidence of my claim, I provided it.

If you choose not to accept it, then that's on you.

If theory does not match observation, then theory is wrong. Richard Feynman said that. It's good advice to live by.

For you to prove your case, you need to prove either A) that autonomous vehicle development is fantasy, & thus not real. Or, B) The problems associated with the technology are unsolvable or insurmountable.

You've already lost on both counts. To summarize: For you to win teh interwebz, you must prove a negative.

If you succeed on your quest it will surly show you truly have l33t skills, & thusly will be awarded 100 of the finest innernetz.

I-Like-To-Bike

The bottom line is that AV's being tested at present are dumb as rocks and lots of money thrown at them and wishful thinking will not necessarily make them smart enough to safely transport passengers too far from driving 25 mph around the outskirts of Phoenix.

Lack of any serious state or local government oversight, as well as test passengers being made to sign NDA's as in AZ may allow the promoters to get away with a lot of hyperbole about their "proven" safety and/or passenger satisfaction with the herky-jerky nature of the dumb ride provided by their unintelligent AI drivers.

wphamilton

I'd mostly agree with you, except for the bolded. When you include the latency of sensors, processing the recognition, decision and mechanical response time I am NOT convinced that the AV reaction is necessarily faster than human. Generally speaking, "on average", humans are slow and they can take from .8 seconds to as much as 1.6 seconds to react to something unexpected in the road. I'll grant that, and in fact it drives me crazy when driving, so for most, or average, human drivers the AV reacts faster.

But not necessarily, and not for all humans. It's not easy for me to find reliable data on the AV reaction speed, but FWIW this article claims it to be .5 seconds. They're selling a faster sensor, so I don't rule out a little cherry-picking for the number, but a good take-away is that it's in that range. My own reaction speed is quicker than that - even if you don't believe me personally, you should at least acknowledge that some humans can and do react more quickly than in half a second and are therefore at least in the same general range as the machine.

It's just a matter of better technology and faster AI though, a matter of time before the machine is faster reacting than I could ever hope to be. But the real difference, the greater difficulty in my opinion, is that the machines still cannot adapt to traffic patterns to anticipate outlier driver behavior and I don't even see that on the horizon.

rydabent

We should never have self driving cars in cities. Too many variables.

Cyclist0108

Today I got close-passed by a Tesla. Apart from firing off the usual barrage of profanity, I didn't think a whole lot about it, but then later on into the ride, I realized the thing tracked the lane absolutely perfectly, as if I was simply not there, and it then occurred to me it might have been on autopilot, which didn't register my presence, or at least the semi-automated correction didn't work.

Leisesturm

You are wrong. How do I know? Because you are alive to write the above (and incorrect) opinion is all the proof I need. AV's are NOT going to pass cyclists with 3' of clearance. Not now, not ever. If you can't abide that ... ...