Californians get first chance to be run over by a Google robot A number of Google's self-driving car prototypes are expected to leave the Chocolate Factory's test track this summer and head out onto some of the public roads around Mountain View, California. The world's largest ad broker said the robocars had undergone "rigorous testing" to ensure the software and sensors work. However, …
If they remain that fugly the only people who'll want to buy them will be pre-teen schoolgirls. And they can't.
On a more serious note, 25mph max in automatic mode or in any mode? Either way, these will
quickly become mobile road blocks and will probably end up causing more accidents than they prevent due to frustrated drivers trying to pass them.
Okay, puzzled at the down vote! (If you're going to down vote, at least leave a comment so we know why!).
So someone on the Reg really thinks these little short range no frills Google cars, will be bought by joe bloggs for private use? Why would you bother when it would be far cheaper to just hire one when needed?
I would have thought anyone buying an auto driving car for private use, would be more likely to buy something more comfortable, like an updated BMW i3, Toyota iQ, or some other similar car that has a few more home comforts in them. (Obviously updated versions/later models with auto-drive added).
This post has been deleted by its author
Most cars reach 30mph on urban commutes, even if their average speed over the whole journey is just 10mph. City driving is largely start-stop: bursts of 30mph when the light turns green, followed by 0mph when the next light is red.
Cyclists often make the same mistake: assuming that because they can hit 12mph, they'll be faster than car drivers averaging 10mph. But cyclists are also doing 0mph at red lights, so their average journey speed is just 4mph.
"Cyclists often make the same mistake: assuming that because they can hit 12mph, they'll be faster than car drivers averaging 10mph. But cyclists are also doing 0mph at red lights, so their average journey speed is just 4mph."
Cyclists stop at red lights?!? Where do they do this?
Whilst I think autonomous driving cars will not be sold in any great number for maybe 5 years (although given the recent surge in electric vehicle purchases that only 2 years ago were being panned due to range anxiety and lack of national infrastructure, who knows?), once they do become mainstream then they really might be a game changer.
What will be the point of me actually owning an autonomous car? For my normal commute to/from work, I only need a car for 2 hours per day. The rest of the time my current car is stationary, depreciating money and taking up space in works car park that I have to pay for.
Between 8am and 6pm, if the car can drive itself, it could be doing something else. It could ferry children between home and school, then ferry someone to the supermarket and back. If this were the case, why would I purchase a car? if the hire-car company does it's sums right*, I'd just hire the autonomous car for the 2 hours per day that I actually need it (give or take a bit for the car travelling to my pick-up point). I still have the convenience of a car taking me direct from home to work and I don't have to wait for public transport. Everyone is a winner, apart from the motor manufacturers who've suddenly lost a big chunk of their market.
* I'm assuming an autonomous car will be significantly cheaper to hire than a taxi because the autonomous car doesn't need to cover the cost of a human taxi-driver.
I'm assuming an autonomous car will be significantly cheaper to hire than a taxi because the autonomous car doesn't need to cover the cost of a human taxi-driver.
That is an assumption that I have seen used by a number of sources, and may perhaps be a reasonable one to make for a commuter vehicle, but I am not so sure it will be the case and, even if it is, will really drive down private ownership of vehicles. First, many areas that have significant numbers of taxi drivers are sensitive to job loss and may take a protectionist stance on the use of driverless cars as for-rent vehicles, especially in cases of individuals leasing them out during the day. Look for example at where Uber has had legal challenges.
Second, the time when someone is most likely to need a car for rent is going to be during peek driving hours which includes the owner who used it to drive in to work. There will be a smaller demand for these vehicles as rentals than there will be supply, which does tend to drive down prices, but...
Third, while the cost of renting one of these has a good chance of being cheaper than a regular cab, assuming there are no price controls, they will still be more expensive than ownership by necessity. No-one is going to rent at or below cost especially as there is going to be some serious liability issues - who is going to want to risk their personal ride home on making a quick $10 for an unsupervised and potentially un-trackable transaction - will you really want to know what sticky mess on the back seat is when you get your car back in the evening?
I did a quick lookup of the number of cabs in different areas. There are now over 20,000 licensed vehicles on London's roads while New York has over 13,237 cabs and more than 40,000 other for-hire vehicles. On the other hand, there are 2.6 million privately owned cars registered to London residents while around 48% of New Yorkers own cars which puts the number of cars there at over 4 million. Uber has more than 14,000 cars operating in New York and 7,000 in London (Uber claims this will grow to 42,000 by next year), so there is clearly demand, but I would guess that there will be strong pushback in areas where there are established taxi services, precisely where there will be a confluence of available cars for hire and people who might hire them.
I am sure some folks will try this, but that it will turn out to be more hassle than it is worth. My guess is that if you want to make money off self-driving vehicles, then your best bet is to sell the security devices that people will want to install in their cars before handing them over to a stranger site unseen.
To the 5 idiots who downvoted me - if you honestly think people won't overtake a car limited to 25mph then you're either a pensioner or you've never actually driven on a public road. No doubt you'll simply downvote this too since apparently you're unable to post a coherent riposte.
Given that you both questioned and expressed some measure of displeasure with the Great Google view of the highway, it could be bots (and lord knows Google knows bots) or those who are eagerly looking forward to this tech for personal reasons.
And yes, this is El Reg. Questioning downvotes just brings on more downvotes.
Hmm... I'll probably be downvoted for reasoning on the downvotes... oh well... It's Friday.
Oh I'm with you on that one, this place must be full of bloody cyclists today. I really pity the poor sod arriving at a junction and just missing the slot before the mobile chicane and its ensuing procession of purple-faced drivers.
Still, one thing that'll result will be a whole load of pseudonyms for "Google driverless car", all in fluent Anglo-Saxon.
25mph may be typical for small residential streets, but you can't realistically go anywhere without using faster roads. The current design doesn't look like anything I'd want to do over 25mph in though.
If they do go on faster roads, but stay at 25mph, who gets the ticket?
"Former business secretary Vince Cable said the £19 million government-funded driverless car projects underway in Greenwich, Bristol, Milton Keynes and Coventry will contribute to a '£900bn industry by 2025'"
If that's not a misprint or a misquote, then I'd like to see what it's based on because that's an incredible sum of money. I know your government are complete asshats when it comes to anything digital but, really, "£900bn"?
Being a foreign observer, you will forgive me for asking, but what are the chances that his successor will be any better? I personally would not be surprised if Vince Cable's prediction was gotten from a Google evangelist, if not directly from Rachel Whetstone. I'd expect Cable's successor to get his predictions from Whetstone's successor. More of the same in every way...
Does seem a tad hard to swallow that.
A quick search suggests $1.6 trillion value this year for the global car industry, which is roughly £900 billion pounds give or take a few hundred billion. Didn't see numbers for 2025. More likely he was just being a twat and putting an infeasibly large random value on something to make it sound impressive, perhaps based on overheard snippets of unrelated discussions, and made up a load of shit in some feeble attempt at getting it to all hang together.
Maybe '£900bn industry' is referring to 'the car industry' as a whole, not just driverless cars. Maybe what he was trying to say was that by 2025 he expects driverless cars to be contributing to an industry that this year amounts to £900bn? And he's just inept in putting it across, or it's been misquoted or taken out of context.
They need to be only let out on the road, if you have a chap with a red flag in front of it. If they actually get to 25 MPH, that would certainly only be a very limited subset of people. I believe that even Usain Bolt would eventually succumb, and become an involuntary speed bump.
Ah! You've hit the nail on the head there. Like in days of old, 4mph & bloke with red flag: what we do is tether a cyclist to the front, with bungee for a laugh. The google car then follows the pushbike. Less tech and when the car stops the cyclist can't break the lights. Well, not for long if we go with the bungee.
what they STILL arnet on the roads?? how many millions of articles have we had about these things being "coming soon" ? what wa that one a couple of days ago - millions of miles / 4 crashes ? I thouht that meant they were on the road - that whole 4 craches figure is even more pointless now i know it was on a test track . I could have done that with my scalextric set!
When these cars are on the road , and by road i mean the one near me that has a boozer on it , i'm not interested
" cars with no one inside"
Nooo, no, no, no. There's the "safety driver", a superfluous meatsack cargo in everyone.
Presumably this will create a new sport amongst the yobs of California, in which the objective is to cause the Googlecar to either go off the road, or screech to a near instant halt, by spoofing the sensors intro thinking that a collision is imminent,
Do in on a bike or in a car, and the angry meatsack won't be able to chase you because you'll be able to outrun his 25 mph limit.
They will have to cruise round forever. No one has yet mentioned how these vehicles will read and understand the Byzantine parking restrictions that can be found on the streets. I don't know of any electronic system cataloguing these. Double yellow lines faded and worn or of varying hues and contrasts as a result of patchwork road repairs, signs flipped round at all angles, 2pm to 4pm anti-commuter restrictions... how does the car know if you plan to be there for a minute or a day? Red routes, corners, schools, zig zags, curb pips, double parking, solid white, people's driveways, boxing in, car club only, electric only, disabled only, bus stop restrictions, mobile library on Thursdays only, yellow plastic boards for temporary restrictions overriding regular plates, permit holders only, 1 hour limits, no returns. On country or seaside roads there's a myriad of variables to consider like soft verge parking, cliff edges, sand traps, drifted sand or snow obscuring markings. Then in the city parking so your passenger can get out without smacking a lamp post or a parking meter or a comms cabinet.
There will never be robots capable of understanding all this. Thus the self driving car will have to have manual controls, will always need a driver, will probably need designated bays with assistive in road technology. Maybe it will work in California where everyone who can afford one has a driveway and a garage and a reserved parking place at work. And I can see a use in controlled environments like theme parks or mega malls or airports as a taxi service but in every day use as a replacement for what we have now? Forget it.
I bookmarked this note from someone who regularly shares the road with the Google vehicles.
Comments like "It's rare to see a self driving car drive slowly." and "I'd say the cars are generally better drivers than their human operators." hardly tally with some of the opinions expressed here.
A couple of weeks ago, I was driving east toward Palm Springs, Ca. While I was making a transition from one freeway to another, I noticed something odd about the car just in front of me. After it finished merging, it moved smoothly 3 lanes over, and entered the carpool lane, moving right at the speed limit. I sped up a bit and got a nice look at the Delphi self-driving vehicle sign on the door, and the two guys inside. The fellow in the driver's seat was watching the road ahead, but his hands were not on the wheel, and the other passenger was a bit hunched over, fondling his phone.
This vehicle was definitely driving itself and going about 60 MPH during all this. And yes, on a very public highway. I would classify the traffic as moderately heavy - moving at or above the speed limit with several car-lengths between vehicles. The transition merge section was also on a curved section of the freeway.
I just wish I had had a passenger, or perhaps a Zaphod modification, so that I could have taken some pictures without jeopardizing anyone's safety...
I live in a resort community. Many folks on the road are from out of town. They don't know where they are, don't know where they're going, and believe this is just cause for ignoring traffic laws entirely (right turns from left lanes, going half the posted speed limit in the passing lane, and so on).
I don't suppose such self-absorbed folks would ever use a driverless car simply because it's easy (cabs have already been invented, after all), but I would love to see them prosecuted to the full extent of the law with the aid of the electronic logs from robot cars.
...that the lawsuits for these rushed-to-the-streets unsafe AVs are in the hundreds of millions per incident for damages. There is absolutely no justification for allowing these not ready for prime time vehicles on the roadways at this time in their development. Google has reluctantly admitted that they have had eleven accidents so far with autonomous vehicles of which at least two were due to the vehicle not the operator.
Small city down town or midtown NYC as few as 3% of the cars on the street could measure 99% of all driving behavior. Instruments for measuring and recording thousands of objects on city streets and sidewalks thousands of times a second in a thousand cars will not swamp Google's big data recording and analysis capabilities. Far from it. A thousand autonomous vehicles in a three mile radius of Grand Central Station might bring order to the streets and sidewalks of New York.
People with 1/3 or more of a brain will downshift and pass these vehicles, probably at high speeds. Some of the more dickly, frustrated subset of the driving population may actively screw with these robot cars out of boredom like the asshats that take a ball bat to mailboxes on a drunken evening.
Really most folks will react just the same as fleeing from a wildcard 90-year-old driver or teen driver glued to their phone that you just can't anticipate what they will do or if they are in full possession of their sensorium. Hopefully Google has tested for this scenario and the cars won't stop, veer, or BSOD when someone "taps" them or blows by them at 120MPH.
Trained dogs could be the best option as safety drivers. Some of the brighter trainees at the guide dogs for the blind school could be streamed into recognising unsafe traffic and consequently using the brake override.
After all, they are able to recognise safe / unsafe driving conditions when escorting their blind owners, and are trained to endure extended periods of concentration, so theres no great conceptual jump in the training.
And I'm sure the doggies would enjoy being in the warm an dry driving seat all day - especially if there was a bowl of biscuits available for every crash avoided
The Register 
Biting the hand that feeds IT
Copyright. All rights reserved © 1998–2025
