Robocar testing in Silicon Valley is underway. Wired reports GM’S Robocar Service Drives Employees Around SF for Free.
Cruise Anywhere, which launched Tuesday, works just like Uber or Lyft: Open the app, type in your location and destination, and wait for your car. It’s available to 10 percent of Cruise’s 250 employees, between 7 am and 11 pm, and uses the company’s fleet of 461 Chevrolet Bolt EVs. (There will still be a safety driver sitting up front, no doubt thrilled their colleagues are headed to happy hour while they work the late shift.)
The program’s launch marks a turning point for Cruise, which GM bought in 2015 for $600 million, as it switches from developing autonomous technology to building a self-driving taxi service. And of course, it’s more than a fringe benefit for hardworking engineers. It marks a new chapter in the development of technology Cruise has spent three years on. Real-world testing will reveal how the vehicles handles the road, and the people it carries.
Of course, Cruise employees aren’t civilians, or a representative sample of a potential customer base. But they are far more likely to keep quiet about hiccups with the tech, and will provide valuable feedback without exposing Cruise to a PR maelstrom if something goes awry.
It remains to be seen how long this program will run, or when Cruise might expand it to the general public, either in California or a state with less restrictive regulations. And it’s not clear how GM’s deal with Lyft to work together on a self-driving car network fits in with Cruise’s decision to build its own app for connecting riders with cars.
But for now, we can say GM’s bid to stay relevant in the age of autonomy just got a boost.
Cruise Anywhere
I have to say that is one ugly car. And it’s not ready for prime time, yet. Those side attachments look flimsy. I suspect a shopping cart might take one out.
That said, a quick check of my calendar shows this is still 2017. So what’s in store?
Automotive Promises Timeline
- Ford: Truly self-driving vehicles by 2021
- GM: Unspecified, with rumors of self-driving vehicles by 2018
- Honda: Self-driving on the highway by 2020
- Toyota: Self-driving on the highway by 2020
- Renault-Nissan: 2020 for autonomous cars in urban conditions, 2025 for truly driverless cars
- Volvo: Self-driving on the highway by 2021
- Hyundai: Highway by 2020, urban driving by 2030
- Daimler: Nearly fully autonomous by early 2020s
- Fiat-Chrysler: CEO expects there to be some self-driving vehicles on the road by 2021
- BMW: Fully self-driving vehicles possible by 2021
- Tesla: End of 2017
Realistic Promises
The above Self-Driving Car Timeline from the June 4, 2017, VentureBeat.
Other than Tesla, which likes to overpromise and under-deliver, the timeline promises are very realistic. It’s easy to see who is way behind, notably Fiat-Chrysler.
The key date appears to be 2020, for highway driving. Trucks will be ready by that date as well.
Regulation is the only missing ingredient. I expect that will be worked out in the next 2-3 years. Then within two years of regulatory approval, there will be a huge rush to driverless trucks with mass adoption coming quickly.
Claims of truck thefts are silly. Within minutes of a truck not being where it is supposed to be, state troopers will be notified. There will not be time for thieves to break into a truck, unload the goods (into what but another easily spotted truck?) and get away. The theft notion is absurd, but it keeps cropping up in comments.
Claims the public will not accept the movement are also nonsense. Trucks will be on the highways, and driverless airport taxis soon follow. There will be higher charges (way higher) for taxis with drivers.
Aging boomers with night vision problems and millennials will be among the first to adopt.
People like me, who like to speed, will be among the last to adopt. Alas, rising insurance costs will eventually force everyone to let the car be in control of driving.
Mike “Mish” Shedlock
Get back to me when there’s a major rollout of 300,000 self-driving vehicles in a major US city. I the meantime I won’t hold my breath. Btw, I never see an estimate of the consumer sticker price for one of these autonomous cars. How much will one of these futuristic vehicles set us back? Will the ordinary consumer and wage earner be able to afford one?
Oh, btw, a “safety driver sitting up front” is cheating and defeats the whole purpose and intent of the autonomous vehicle. If you want us to take self-driving cars seriously that has to stop forthwith.
Ordinary consumers already cannot afford cars. They are about to become fulltime pedestrians who can occasionally afford a rideshare service.
In my neck of the woods one can buy a nice used normal car for $7-$10,000. That’s affordable for most. In fact, most poor people in America own cars. Even illegal aliens can afford one. In California they signed up nearly a million illegals for driver’s licenses. Why would they need a DL if they can’t afford a car?
You avoided answering my question: How much would it cost a consumer to buy a self-driving car with all it’s bells and whistles? I never see that discussed by the pro-autonomous car crowd. Why?
Probably because the cost is not yet known. When it is there would be no point in keeping it a secret.
Arlington Texas has started with two 12 passenger self driving shuttles on a six month lease, estimating a cost of $272,000 over two years. Seems totally affordable to me, especially given it’s not their money. Of course the are spending LOTS of other people’s money that they have yet to “collect” on the new Cowboys stadium and the new baseball park to replace the old baseball park (yet to be paid for) and the rest of their sports heaven utopia.
I’m going to get me one…once my “grant” comes through.
Buy? BUY??? The article says they get them free!
It just has to be :
http://www.miamiherald.com/news/business/article166646987.html
https://www.reuters.com/article/us-germany-emissions-quotas-idUSKBN1AR0XL
I’ve figured price in the showroom of $3000. It’ll probably be more ’cause the market will bear more. And, there are good odds that one of these car companies will have a clearly superior version-one product. That car company will be swamped with orders, so they can crank their price up. Think of how many car/truck buyers would not blink at level 5 for ten grand.
That said, you can do your own arithmetic pretty easily.
In 2009/2010 the dev vehicles were around 300k for the whole car. The extreme bulk of that cost was tech stuff that follows Moore’s Law. So, 2015: $30k. 2020: $3K. 2025: $300. Build cost.
Two things make that calculation very, very rough:
1) Dev vehicles are custom built in onesee, twosee quantities. They use inherently expensive equipment because they are optimized to save expensive developer time. (E.g. Right now, if you’re building a car for dev purposes, you’re probably buying cameras at $2500 a pop. Pretty much the same fundamental camera as you can get for $5 from one of the electronics distributors. The $2495 buys a bunch of packaging and hardware/software wrappers, easily saving a few days of dev engineering time, so it’s money well spent. Another example: Mish’s picture shows a car with two LIDARs on it. Base price, per LIDAR: $7500++++. The main LIDAR manufacturer is in a bind. They can’t ramp up production ’cause a few dozen dev outfits won’t buy diddly. The manufacturer says as soon as the orders for production quantities come in, the price cuts to 1/10th. WayMo, meanwhile, sues Uber ’cause they figure Uber stole WayMo’s design for a right-now, under-a-thou LIDAR.)
2) Full, level 5, self-driving doesn’t exist. So, the people building a driver can only make a rough guess about what kind of processing power version one of level 5 will take. Put crudely, how many GPUs does level 5 need? Who knows. And, meanwhile, all over the world, people are falling all over themselves to build custom neural net processors. An algorithm hack or a custom chip can turn a $5,000 processor box in the truck in to a $50 module under the dash. But car designers can’t count of such a hack coming in the next couple, three years, so …
Anyway, car companies aren’t throwing billion dollar bills at the tech without considering end user cost.
The technology required to build a Google self-driving car alone runs $150,000. Under the best of circumstances that might get reduced by perhaps 50% with mass production and advances in technology – leaving the consumer with a $75,000 price tag – which would be well beyond the range of an average income earner.
Americans simply won’t tolerate being forced to use public transportation or to share a vehicle with 4 other people. Major inconveniences that won’t be tolerated.
Cost is only one obstacle. There are a myriad of unintended consequences that have been discussed in detail on this board that will sabotage any effort to launch autonomous cars into the mainstream for the next 25 years.
But keep dreaming. It’s a favorite American pastime.
LFOT: You’ll never see a 300,000 roll-out for a city. Doesn’t make sense. If you can make 300k items, why not flood the world with 300,000,000?
Driver up front: These are development vehicles. They exist to find flaws and identify improvements. When the driver isn’t wanted, the fat lady will have sung. What you’re waiting for is when the AV test cars disappear from where they are now. That’ll mean that environment is “solved” and the developers have moved on.
It would be nice if there were more info about how testing is going in India. India has about every driving situation there is right on their doorstep, right? If the Indians don’t dominate self-drive they should just shoot themselves to get out of their misery. 🙂
good grief. luddite much?
Mish,
Out of your list of car manufacturers…
How many will even be around by the time stamp of their predictions?
Likely all of them
The question is: how many will exist by 2025, and in what capacity?
Mish,
a more precise way of describing the goals:
Ford: Truly self-driving vehicles by 2021 ( level 5)
GM: Unspecified, with rumors of self-driving vehicles by 2018 (level 4-)
Honda: Self-driving on the highway by 2020 ( level 3.99)
Toyota: Self-driving on the highway by 2020 ( level 3.99)
Renault-Nissan: 2020 for autonomous cars in urban conditions (level 4) 2025 for truly driverless cars ( level 5)
Volvo: Self-driving on the highway by 2021 ( level 3.99)
Hyundai: Highway by 2020, urban driving by 2030 ( level 4)
Daimler: Nearly fully autonomous by early 2020s (level 4.75)
Fiat-Chrysler: CEO expects there to be some self-driving vehicles on the road by 2021 ( level 4?)
BMW: Fully self-driving vehicles possible by 2021 ( level 5?)
Tesla: End of 2017 ( level 3.75)
How about the Beach Boys’ ‘409’?
Autonomous cars cannot share the road with idiots who do not(care to) know the rules of the road. In Anaheim, CA all the hispanics yield during right turns to oncoming left turners and conversely think that left turners have the right of way, never yield to pedestrians, never stop at a red light or stop sign when making rights, pass & speed on the left, squat in left lanes, do not understand what merging traffic is nor to move over to accomodate them, rarely bother to stop at a metered on-ramp, and generally have a me first, I’m crossing the border mentality. Not just hispanics but people of all backgrounds are crap drivers. How are computerized cars who follow a set of rules(code) going to adapt to such idiocy? Without ALL cars being autonomous there can be no sharing of the roads by man & software.
You are correct of course as we saw horses give way to vehicles a hundred years ago. The thing is that the cost of these mandatory vehicles will be too high, forcing most people to abandon private ownership and choose Uber type transport instead.
This fits into the progressive theme quite well, limiting where and how we travel and prompting even greater living density. Agenda 21 at play I think.
Freedom is just another subversive plot.
I suggest you try following all the traffic rules and see how far you get before you have to violate one. Some examples:
– A Mail van has stopped to put mail into someone’s mail box. But the center line is a double line so you are not allowed to pass it – are you going to stay stuck behind the mail van for miles?
– Similar to the Mail van case – but a garbage truck. The garbage truck stops and a garbage man waves you past – but you would have to cross the solid center line to do it.
– There’s a non-stop stream of people crossing a marked pedestrian crossing in front of you. Do you sit there forever? Or do you wait for slight gap to appear and slowly and carefully push through (illegal)?
– You want to turn left but there is oncoming traffic so you sit in the middle of the junction and wait. The light goes red but you are still in the junction when the last oncoming car stops. You have broken the law (you are not supposed to even enter the junction unless your exit is clear – but if you actually did that you would never get anywhere)
– The traffic lights have broken in a multi-lane junction. You are supposed to treat it like a 4-way stop – but there are multiple lanes in each direction and lots of traffic. You cannot know whose turn is next, so you will almost certainly just ease out when you think it’s about your turn and break the rule.
– You arrive second at a 4-way stop, but the other guy who arrived first doesn’t move and just sits there (thinking you were there first, perhaps). Maybe she even waves you through. So you go – and break the law.
These are just a few examples off the top of my head, but there are many, many cases where you need to break the traffic rules in order to make progress in a messy real-world environment.
Don’t confuse them with the facts.
Reblogged this on World4Justice : NOW! Lobby Forum..
These cars have been ‘testing’ within the neighborhood that I reside for months. On an average day I see at least 3 to 4, there is a quite a presence. It’s been fun watching the cars react when letting the dog wander a bit into the road as the vehicle passes. (they come to a complete stop in case you’re wondering)
It’s 2017
And we are talking city
The timelines are quite believable even if what you say is true (which could also be GM needs to catch up)
No one doubts your enthusiasm!
https://www.youtube.com/watch?v=XHIq0MMViZg
Shows some of the current weaknesses.
Drives like an old lady.
2:2x for old-lady hesitation.
2:37 for an animal.
Strengths, too. Does will with lots of walkers, bikers, buses, delivery vans, and the usual sloppy drivers.
This is from the outfit GM bought.
yup – appears safe and sound
driving the speed limit is desireable BTW
Remember – only the successful cases make it to Youtube. If it fails 1 in 100 times (and I suspect it is a lot worse than that) that video is quietly shelved.
Autonomous vehicles will dominate the roads eventually, just not as quickly as some believe. Though there will be some early adopters, there will also be a lot of resistance from most individuals who prefer to be in control of their vehicle. In addition, the cost will be prohibitive at first (my personal estimate is 50k – 150k depending on the model).
Early adopters will include the taxi, and long haul trucking industries. Individuals who cannot drive anymore (due to various disabilities) will be early adopters, if they can afford it. Some individuals will be early adopters because they like their toys.
I would estimate that mass adoption will occur around 2030, when a basic autonomous vehicle will drop in price to perhaps 30k, and insurance costs will become too expensive for non-autonomous vehicles.
Those who cannot afford an autonomous vehicle on their own, might share one with a group of friends or family, or simply use services like uber or left.
I love the notion of “mass adoption” like it is all voluntary. What we do in the future is determined by our programming. We like what we like because we are SUPPOSED TO.
Tell me what Facebook thinks.
Tell me what Google thinks.
Tell me what Bezos thinks.
Mass adoption happens when the cost/benefit is too positive to resist. Just like when cars were “mass adopted” and horses stopped being used for basic transportation. Only the paranoid believe it’s a conspiracy.
I fail to see when/where/how a self-driving cars offer the irresistible positive cost/benefit calculation that you assume here…
will SD autos be cheaper? doubtful.
will SD autos be measurably faster/better? no.
I can see logistics/trucking/distribution co’s investing in SD trucks to dramatically cut payroll/costs, but individual drivers/households simply won’t have any incentive to pony up for SD auto tech.
SD autos for the masses is a dream that will crash for lack of a market.
Hey mp. Though you may be correct, I doubt it. People said the same thing about trains, planes, and automobiles when they were first introduced. Same for ATMs, smartphones, self-checkout, etc.
Here are just a few of the cost/benefits I’m talking about:
If you replace 2 or 3 vehicles that you typically use 2% of the time with one autonomous vehicle, plus the occasional Uber, that is a huge cost/benefit. Why have $100k tied up in your driveway when you can make it $50k.
If you only have to insure one vehicle (rather than 2 or 3), and the insurance cost on that vehicle drops by 80%, that is another big cost/benefit.
If you only have to maintain one vehicle (and it can drive itself to the service center when you don’t need it) that is another cost/benefit.
When you can get in your vehicle and have it drive you safely while you make phone calls, respond to emails, take a nap, etc that is a positive cost/benefit.
It would be easy to add a hundred more of these, but I hope you can see my point.
Now, I could be wrong about the cost/benefit appeal, but I feel pretty confident that I’m correct. Where my crystal ball is hazy is not “if”, but “when”. My estimate is around 2030.
my company’s CTO recently stated, in 10 years he expects not to need a personal vehicle anymore as there will be a fleet of level 5 fully autonomous vehicles ready to take him anywhere anytime. in my company’s business, which is in direct technological support of the insurance industry, he has to know what the automotive trends will be.
Speaking of insurance, I’d like Mish to explain why he thinks insurance costs will rise.
If robots make the roads safer, it should be safer for everyone. That would mean accidents and insurance rates would go down. Not up.
Insurance will rise for those who insist on taking the wheel themselves!
only if the actuarial data supports this conclusion.
and if SDC’s turn out to be more hazardous than blindfolded teenage drivers? the insurance on SDC’s will skyrocket.
SDC’s are going to be targets… abused to the maximum. i doubt that insurance rates on SDC’s will fall below the median.
It will be safer for everyone but there will still be some accidents. As a crude estimation, since 97% of accidents are currently human error, and you assume that continues in the future, then human driven vehicles should pay 97% of all insurance premiums, while autonomous vehicles should pay the other 3%. As the number of human driven vehicles declines, they will have to pay higher and higher premiums. Eventually, they will be too high to make those vehicles practical. It’s not a conspiracy, it’s just a logical conclusion.
“97% of accidents are human error” – data based on 100% human drivers.
when the roads are 25% SDCs will 97% of accidents still be human error? YOU CANNOT MAKE THAT ASSUMPTION w/o DATA as the denominator is significantly different in the new environment.
and when the roads are 95% SDCs (as you predict) will 97% of accidents still be human error? LOL.
It’s going to have to be a HUGE fleet, because everyone wants to go to/from work at exactly the same time.
Self driving cars would be as popular as remotely piloted sex.
Our decisions are determined by the choices we are afforded.
Multiple choice:
A) self driving car or self driving Uber
B) incredibly expensive and libelous human driven models.
C) Bus
D) Walk
Throw cost/convenience in and you are on target. There will be the B) holdouts – and I’m fine with that if they can afford the insurance if their driving performance is worse than average (if they are better than average, their insurance can go down, and with a significant percentage of cars equipped with 360 cameras, we will know a lot of the time who was at fault, speeding, etc.)
I’m trying to model the “on demand car” vs. Bus option – and again all I come up with is cost vs. convenience. However if you have a “get me from point A to point B” app on your phone, and it says “walk 50 yrds and wait 1 minute for the bus, then walk 100 yds at the other end – total cost/time $2/12 mins” vs “wait 6 mins for car – total cost/time $8/9 mins” it comes down to consumer choice.
jump on bike ride in 15 minutes at zero cost
It wasn’t a choice – but my primary mode of transport is my bike. It started as a fitness thing, I’m not trying to save the World or anything, but I just started asking myself the question every time I went into the garage “Why can’t I ride my bike to do this?”
An old saying from my Bro-in-Law in Scotland: “Money can’t buy you happiness, but it can buy you a bike, and that is pretty close.”
Notice in that photo how all the other cars are hanging back?
Just like a cop car where everyone suddenly drives within the speed limit and leave a safe distance from it, except here they’re just keeping a safe distance…..
https://image.ibb.co/jF0MPv/sketch_1502494649212.png
I read an article last week that certain auto companies are joining together on electric vehicle and driverless vehicle research. This points to one thing in my opinion, the results might not be worth the cost and they are sharing startup costs to guard against a downside if this technology doesn’t provide financial returns expected. Smart move since this would not be the first big thing that fell flat from drawing board to public use. Someday this technology will have to stand on its own without government subsidies. Great example is tesla.
Tesla. Exactly. Government subsidized wet dream that has yet to make a market dime, dependent upon billions of taxpayer money, and valued at more than Ford.
THAT is our self driving car.
Couple things bring these guys together:
1) With connected cars, especially, the biggest player wins all the chips. Do you want a big slice of nothing or a small slice of a huge pie? Same network effect will be true for controls. Remember when you avoided renting a GM car ’cause their controls weren’t exactly like the Japanese cars you knew how to drive? Sharing software encourages the procucts to be compatible.
2) Things like Baidu’s Android-like play. Baidu is making the software available so smaller Chinese outfits can clobber the big, established players by swarming over them.
Let’s see how well the self-driving car does when the world gets messy, as it often does. Policeman directing traffic, hole in the road, tree across the road, debris in road, flooding, faulty traffic signal, etc, etc. If it can’t understand something unusual – and it can’t – it will either just sit there (passengers will have to get out and walk), or it will likely do the wrong thing. This is always the problem with AI – no understanding of the real world and no flexibility to deal with the unexpected.
I own and operate CNC equipment which is far more mature technology than self driving cars and not nearly as complex…..and those suckers go cray cray from time to time. Sometimes a reboot does it and sometimes a tech from Italy at a cost of thousands of dollars.
I
CAN’T
WAIT
Currently many of us own multiple vehicles to satisfy several purposes. In the future we will need multiples…for backup.
short sighted. there won’t be a need for police to direct fully autonomous vehicles.
If you want us to take self-driving cars seriously that has to stop forthwith
So I can’t ride my harley anymore cause of insurance? What about my hot rods and some of us like to drive? One more reason to stock up on guns and ammo. Some of you talk like this is a given. It’s a given its going to be a disaster. I’m glad I’m on the downside of life expectancy as everyday I feel as if I’m surrounded by one more moron.
You can still drive all your vehicles, as long as you are able to afford the expense (just like now). Just be aware that maintenance and insurance costs will creep up over time. And that parking spots will start to disappear over time as well. Again, it’s not a conspiracy. It’s just a logical conclusion of the coming changes.
By the way, how will stocking up on guns and ammo help you to drive? I am continuously amazed at some Americans; angry, and willing to resort to violence when they don’t grasp reality! Like Mr noble, I feel there are too many morons in America!
As diogenes said, cars are already getting to be too expensive for most. A shared fleet of autonomous vehicles is pretty much the only system that will work for all. That will be the endgame assuming worldwide industrial collapse doesn’t arrive first, which is unlikely. Enjoy driving your own vehicle while you still can.
I wonder how well they will work when driving internationally. I’ve driven through the mountains of Mexico where the roads are barely discernible ruts that go through streams and shallow rivers – will an automated car simply refuse to go where most trucks can go? Or through crowded city streets of mixed traffic (cars, bikes, pedestrians, donkeys, motorcycles) ignoring all traffic rules where hitting something means jail time – will the car company be liable when (not if) that happens and send someone to sit in jail in your stead? Will it be smart enough to recognize a military checkpoint, where running the checkpoint means being gunned down? Or when locals put up their own “toll both” (rope across the road)? Screw American roads – I want to see this thing drive itself the length of the Pan American Highway without getting the passengers jailed or shot.
Or will we be required to travel only pre-planned routes, approved for our safety, caged in our own country? From what I can tell, this is what most people want – to be freed from the freedom to travel.
You are right, there will be scenarios that stump AI cars, but then there are many people who can’t reliably drive “home to grocery store and back” where I live – even the sober ones.
So if 95%+ of the trips are within the capabilities of your AI car or service, you can focus on the others yourself.
For me and my needs the AI cars I see toddling around my HQ office area seem to be doing fine – when I even notice them.
“Those side attachments look flimsy. I suspect a shopping cart might take one out.”
They grow back!
Up in Northern Michigan, the auto industry just had a conference. GM had one of their autonomous cars drive up from Detroit, all but 18 miles of the 220 miles were autonomous. Michigan has legalized autonomous vehicles everywhere in the state.
Where is all this going? We’re headed for a massive disruption to the transportation industry. Fleets of autonomous electric vehicles will be managed by a few companies, perhaps GM and Ford will transition into this business. There will be little reason to actually own a car if you can hail one quickly. No car, no parking, no insurance, no fuel, no maintenance – this will be financially compelling for many people. This will also impact the petroleum industry as demand for fuel will drop by 33% to 50%. And parking lots will be re-purposed. These changes will happen fast, my guess is around 2030.
I’m sure someone has brought this up this issue, but I seem to have missed it: Don’t police departments get a huge portion of their income from traffic tickets and from seizing the cars of people whom they have accused (but not convicted) of crimes? If self-driving cars become common, there will be far fewer people to ticket, since the cars will presumably not speed, and passengers can be drunk without it being a problem – no more $5,000 DUI fines for the police to take in. And if the cars being ridden are not owned by (for instance) accused drug dealers, seizing the car will require going up against the car-owning company, rather than a powerless individual. If police departments lose large amounts of their funding, things could get interesting.
The “Self-Driving Car” is Only an Oxymoron
Over at Tesla, Google, and Uber — and now the contagion has reached Ford, General Motors, Chrysler and beyond — the smartest guys in the room are talking autonomous vehicles. Over at every hedge fund, venture-capital and wealth-management shop in the universe, the smartest guys in the room are throwing money at the concept. Why? Because it’s the Next Big Thing, that’s why. Billions of dollars are in play.
Which is why we are seeing an avalanche of faux-news stories about the coming era of driverless cars, how they’re on the streets now, how well they are doing in testing, how soon there will be nothing but driverless cars on all our roads. And all this chum in the financial water has served its purposes: the hedge fund sharks, and the Masters of the Universe they serve, are in a feeding frenzy; and the gullible public is giddy with anticipation.
Meanwhile people with a lick of common sense are saying, wait a minute, doesn’t this sound oxymoronic, like clean coal, or safe sex? In today’s world, people with licks of common sense do not get funding to answer their questions, and therefor the skeptical questions you might have about “driverless cars” are almost unanswerable. Until right now, right here:
Is there such a thing as a “driverless car?” Not yet, there isn’t. The conditions for allowing “driverless cars” on the public roads in a few states unanimously specify that the driverless car has to have a driver who is ready to instantly take control of the vehicle. Moreover, what they are driving and testing are prototypes and jury-rigs; no one has yet built an autonomous vehicle. (Tesla cars offer “auto-pilot,” but it isn’t.) So almost all the stories you have read and seen about “driverless cars” on the road are fake (some fastidious journalists write about testing cars that are capable of becoming autonomous, but most people read right through the fastidiousness).
How are the potentially driverless cars doing in their testing? Awful. For example, in the first week of March, Uber’s 43 test cars in three states logged some 20,000 miles on public roads. Their drivers had to intervene and take control away from the software, an average of once every mile. Critical interventions, required to save lives and property, were counted separately; they occurred every 200 miles. Which makes your life expectancy, as a passenger in a truly autonomous car, approximately four hours.
How much will a driverless car cost? No one has a clue, because no one has built one yet. As Consumer Watchdog put it in its devastating report on the imaginary industry, “No completely self-driving vehicle is offered for sale today, and notwithstanding a great deal of marketing hype, no manufacturer has set a firm date when it will market a passenger vehicle that is able to operate in all conditions without human intervention, or, importantly, what it will cost to buy.” Just to hazard a guess here, my bet is that the early adapters to driverless cars are going to have to choose between buying the car and buying that second fully-staffed luxury yacht.
Would driverless cars need new infrastructure? You bet your booties. No one has even begun to plan, let alone estimate the cost of, a national network that would support the car-to-car, car-to-road and car-to-satellite communications that would be required by tens of millions of driverless cars. Figure that out, and then specify who is going to pay for the staggering cost, while our bridges collapse, our interstate highways continue to rot, and our potholes continue to eat 18-wheelers. Then you’ll have a plan.
What about insurance? Worry not, the insurance companies are on this, 24/7, with a full-court press in every state. Not, as you might think, to figure out how to equitably insure the vehicles and arbitrate liability; but to demand that all the state immediately suspend regulation of and restrictions on the insurance companies lest — and this is a quote from a top insurance lobbyist — the regulators “chill this promising technology and the huge advances in overall public safety it promises.” Speaking of chill — did you just feel one? You should have.
Bottom line: “driverless cars” are not here, and not coming. Like artificial intelligence, virtual reality, genetic engineering and other “next-big-thing” oxymorons, what we’re really talking about here is a high-tech con, designed to separate real morons from their money.
http://www.dailyimpact.net/2017/08/03/the-self-driving-car-is-only-an-oxymoron/