tyrion

I predict driverless cars will start to be in actual use transporting passengers (not in a of dev/testing context) in 2018. How prevalent they are will depend on politics. Some governments will be more cautious than others.

I-Like-To-Bike

And keep cheering on the Silicon Valley practice of beta-testing their unproven prototypes on the consumer, though this time it is being conducted on the general public on the highway rather than on its personal computers and devices. This time the defects and shortcomings of the prototypes will not be limited to "crashes" on desktop computers.

You may be looking forward to reading about the results when the beta prototypes are tested under more difficult conditions, like driving at night, or in rain and snow, or bridges or tunnels, busy city streets, or tree shade dappled highways or the multitude of different conditions that humans handle everyday. But you may be waiting quite awhile.

There is a reason why Waymo picked dry and sunny Arizona for their beta testing on public roads and not the dry and sunny and far more crowded highways in your backyard. California has regulations to prohibit the disregard for public safety involved in testing such unproven vehicles on public roads. Arizona legislators apparently have less concern for the safety of its residents.

mr_bill

Uh, California regulates autonomous vehicles. So does Arizona. So does Pennsylvania. So does Michigan. So does Massachusetts. So does New York State and New York City. So does.... Yet autonomous vehicles are deployed (or in the case of NYC, soon to be deployed) in all these places, because they are dry and sunny and uncrowded?

Do you have a pointer to the memo that you've obtained from Waymo about the secret reasons why they test in Arizona?

-mr. bill

Leisesturm

This bothers you why? Part of the problem as I see it, is that (IMO) non-essential motor traffic should be totally absent from the urban core of any city larger than ~100K. Seriously. But IF the present paradigm of allowing motor traffic to operate unimpeded in busy downtown areas continues they SHOULD always defer to pedestrians. I live in an area with a lot of uncontrolled intersections and I have to wait until ALL traffic has cleared in both directions before I can cross an intersection safely. Sometimes this takes a few seconds, sometimes it is several minutes. Sometimes a driver gives me a break. It is totally up to them. The power is all in the motor traffic. That (IMO) should not be the case in urban cores and, in fact, there are statutes that give right of way to peds and cyclists in my very locale but there is lax enforcement.

After 19 pages of point/counterpoint I see no new information to disabuse this writer of an opinion that believes the answer is an unqualified 'Yes'. A self-driving car world would indeed make it safe for cyclists because self driving cars do not have personal biases against peds/cyclists/other vulnerable road users, that are purely subjective in nature. When basic physical laws of collision avoidance dominate the interactions between large powered road using vehicles and smaller human powered or pedestrian road users the safety of the latter should not remain in question.

Ninety5rpm

Bob Lutz weighs in.

www.autonews.com/article/20171105/INDUSTRY_REDESIGNED/171109944/bob-lutz%3a-kiss-the-good-times-goodbye

noisebeam

Every day thousands of new drivers beta test their skills after passing a minimal alpha test.

Even worse is these new drivers can't pass on their learning to the next new driver so they never get better.

tyrion

Lol. +1

genec

fify

Yeah... but somehow these "beta drivers" don't seem to phase some folks... in spite of the well documented crash rate of something like 3X greater than the general population. Gee, is it any wonder that teen drivers pay more for auto insurance?

I-Like-To-Bike

Nor do they need to stay within the friendly confines of certain selected Google approved/selected routes in the Phoenix, AZ area in order to avoid challenging an inability to operate safely with no one at the wheel in the dark, rain, snow, slippery roads, dense urban traffic or any/every other scenario that doesn't match the sunny day environment selected for these demonstrations.

I suppose the only drivers the overwrought motorist-bashers/Google car fan bois/day dreamers would have are whatever Phase driver is equivalent to a vehicle that stops by the side of the road whenever the sun goes down, or approaches a tunnel, bridge or anything else not meeting the sunny weather, carefully selected Google-friendly dry clean routes of Phoenix, and doesn't budge until road conditions change to meet the severely limited safe operating parameters of the current breed of driver-less motor vehicles.

02Giant 

It is a nice dream for some, but it isn't going to happen where all but autonomous vehicles will be legislated off of the roads.

genec

Apparently you missed this, earlier in the thread...
Just how well do you think your average new teen driver handles that construction zone with the bad light and cyclists riding by.

And regarding night time... you do realize that new drivers are not allowed to drive at night...
Remember the self driving cars only need to be just better than the average motorist to be an improvement... just as airbags and seatbelts don't save all lives... yet, they are mandated.

genec

I suspect you are right... there will still be a need for specialized vehicles... and older vehicles will remain on the roads until they "expire."

But it may become more and more difficult to operate gas vehicles... if electrics become more popular and gas stations become fewer and fewer. No doubt this would be long in the future... just as it took time for cars to replace the horse and buggy.

noisebeam

These Waymo vehicles are geographically restricted to mapped areas when they don't have a human backup driver. They are actively increasing the mapped area using autonomous vehicles with human backup.
They can drive without human at night and light rain.

tyrion

I expect this to be a common procedure, the tested and authorized zones will simply grow over time until they cover almost everything drivable.

Ninety5rpm

Agreed. That's a point on which I disagree with Lutz. But I do think economics and convenience will naturally cause most vehicles on the road to be autonomous within ten years; maybe within five years.

It will happen very, very quickly because the hurdle to switch is so low: just leave your car in the garage and take a driverless Lyft for a great price instead. You can go from less than 1% doing that to 90+% doing that literally overnight. The biggest issue will be in building enough AVs to meet the demand.

And just as quickly, the roads will become far safer for cyclists.

Ninety5rpm

Exactly.

The operable range and operable conditions are at a baseline, and will continually increase until you can take one anywhere and any time a human-driven vehicle can go.

The rate of increase is what I think is hard to grasp. Only one car has to be taught how to drive in a particular area - and suddenly all the cars have the accumulated knowledge.

02Giant 

Expire? What kind vehicles have you owned? There are 90+ year old vehicles still being enjoyed by their owners, how long until they "expire"?

I fail to see a majority of the general population give up their vehicles. For the most part, public transportation is a failure in getting the masses out of their cars. And, I don't see Legislation happening.

02Giant 

Who will build the Lyft system's infrastructure? overnight...

I don't doubt the technology will continue to grow and more autonomous vehicles will end up on the roads. But, it is going to take a long-long time before a majority of vehicles on the roads are autonomous.

Leisesturm

Just like that, eh? Just walk away from a $60K investment... seriously? Some people have over six figures invested. Yah, the mouthbreather driving the beater held together with spit and baling wire... yah, him. The dude with the Veyron, not so much. Better show him the money!

tyrion

That's wildly optimistic. Outside of dense cities the business case for driverless taxis could take decades to develop and displace existing autos.

noisebeam

The number of people who have investments in vehicles is extremely slim. Most depreciate.

Leisesturm

You're splitting hairs. No one is just going to stop driving a perfectly good 3 year old Escalade without some kind of $$$ incentive (or coercion) just because.

Ninety5rpm

They already are in some places, like Manhattan and San Francisco where the relative cost/hassle of driving and parking your own car is much higher than in other places. But the balance will tip in favor of hailing everywhere else once the price drops when you won't have to pay for a driver any more. As soon as that starts to happen the demand for parking will drop and so will the prices, rendering more and more parking garages to be no longer economically viable. As garages close and prices start to increase again trying to find a new equilibrium more and more people will continue to opt for hailing instead of driving their own vehicles.

The good news is you'll be able to buy a perfectly good 3 year old Escalade for a lark.

Now is not the time to buy a new car! LOL!

Ninety5rpm

Maybe. I obviously don't think so.

All the technology already exists for driverless taxis - they're just refining the software. Most new cars can be developed in 18 months; I see no reason why this can't happen with driverless taxis. And there are companies already working on this. Some in the general public might be getting rides in such taxis this year (yeah, within 7 weeks).

I-Like-To-Bike

More news from the Internet:
Driverless Cab Service to Launch in Arizona
by John Lister on November, 8 2017 at 01:11PM EST
https://www.infopackets.com/news/102...launch-arizona
Extract:
"...During the tests, both the driver and passenger seats are empty. A Waymo employee sits in the back of the vehicle and can hit an emergency brake button, but otherwise has no control.

It appears to be the first time driverless vehicles have been tested on public roads at a normal driving speed. In tests by other companies, a human has sat in the driver seat with the ability to take over at the wheel if needed. (Source: startribune.com)

Fair Weather Makes Easier Test Ground

That's partly because other states have tighter controls over autonomous vehicle testing. As well as the looser rules, Waymo picked Arizona for testing because it has little rain, snow or fog, both of which are more of a challenge for current self-driving technology.

The next step will be offering rides to some of the 10,000 people who signed up to be test passengers. In the beginning, a Waymo employee will sit beside passengers in the back, but this will eventually be phased out. If the testing goes to plan, Waymo will offer a taxi service where passengers can take a ride to the destination of their choice without a driver on board.

3D Map is Key to System

To start with, the testing program will only allow journeys that are entirely within a designated zone around Chandler, a suburb of Phoenix. That's because Waymo's technology is based heavily on creating 3D maps of an area. The cars then use this as a reference point, making it easier to distinguish between fixed objects, such as trees, and things that could move such as vehicles or people."