Nj Drivers Test Parallel Parking Size

I have spent quite a bit of time lately thinking about autonomous cars, and I wanted to summarize my current thoughts and predictions. Most people – experts included – seem to think that the transition to driverless vehicles will come slowly over the coming few decades, and that large hurdles exist for widespread adoption. I believe that this is significant underestimation. Autonomous cars will be commonplace by 2025 and have a near monopoly by 2030, and the sweeping change they bring will eclipse every other innovation our society has experienced. They will cause unprecedented job loss and a fundamental restructuring of our economy, solve large portions of our environmental problems, prevent tens of thousands of deaths per year, save millions of hours with increased productivity, and create entire new industries that we cannot even imagine from our current vantage point. The transition is already beginning to happen. Elon Musk, Tesla Motor’s CEO, says that their 2015 models will be able to self-drive 90 percent of the time. And the major automakers aren’t far behind – according to Bloomberg News, GM’s 2017 models will feature “technology that takes control of steering, acceleration and braking at highway speeds of 70 miles per hour or in stop-and-go congested traffic.” Both Google and Tesla predict that fully-autonomous cars – what Musk describes as “true autonomous driving where you could literally get in the car, go to sleep and wake up at your destination” – will be available to the public by 2020.

Make full stops, look left and right a couple times before going after stopping, use your blinkers, wear your seatbelt, have both hands on the wheel and no distractions, make sure the instructor in the car with you is wearing his/her seatbelt, drive the speed limit on the course, and know how to parallel park.

How it will unfold Industry experts think that consumers will be slow to purchase autonomous cars – while this may be true, it is a mistake to assume that this will impede the transition. Morgan Stanley’s research shows that cars are driven just 4% of the time, which is an astonishing waste considering that the average cost of car ownership is nearly $9,000 per year. Next to a house, an automobile is the second most expensive asset that most people will ever buy – it is no surprise that ride sharing services like Uber and car sharing services like Zipcar are quickly gaining popularity as an alternative to car ownership. It is now more economical to use a ride sharing service if you live in a city and drive less than 10,000 miles per year. The impact on private car ownership is enormous: a UC-Berkeley study showed that vehicle ownership among car sharing users was cut in half. The car purchasers of the future will not be you and me – cars will be purchased and operated by ride sharing and car sharing companies. And current research confirms that we would be eager to use autonomous cars if they were available. A full 60% of US adults surveyed stated that they would ride in an autonomous car, and nearly 32% said they would not continue to drive once an autonomous car was available instead.

But no one is more excited than Uber – drivers take home at least 75% of every fare. It came as no surprise when CEO Travis Kalanick recently stated that Uber will eventually replace all of its drivers with self-driving cars. A Columbia University study suggested that with a fleet of just 9,000 autonomous cars, Uber could replace every taxi cab in New York City – passengers would wait an average of 36 seconds for a ride that costs about $0.50 per mile. Such convenience and low cost will make car ownership inconceivable, and autonomous, on-demand taxis – the ‘transportation cloud’ – will quickly become dominant form of transportation – displacing far more than just car ownership, it will take the majority of users away from public transportation as well. With their $41 billion valuation, replacing all 171,000 taxis in the United States is well within the realm of feasibility – at a cost of $25,000 per car, the rollout would cost a mere $4.3 billion. Fallout The effects of the autonomous car movement will be staggering.

PricewaterhouseCoopers predicts that the number of vehicles on the road will be reduced by 99%, estimating that the fleet will fall from 245 million to just 2.4 million vehicles. Disruptive innovation does not take kindly to entrenched competitors – like Blockbuster, Barnes and Noble, Polaroid, and dozens more like them, it is unlikely that major automakers like General Motors, Ford, and Toyota will survive the leap. They are geared to produce millions of cars in dozens of different varieties to cater to individual taste and have far too much overhead to sustain such a dramatic decrease in sales. I think that most will be bankrupt by 2030, while startup automakers like Tesla will thrive on a smaller number of fleet sales to operators like Uber by offering standardized models with fewer options. Ancillary industries such as the $198 billion automobile insurance market, $98 billion automotive finance market, $100 billion parking industry, and the $300 billion automotive aftermarket will collapse as demand for their services evaporates. We will see the obsolescence of rental car companies, public transportation systems, and, good riddance, parking and speeding tickets. But we will see the transformation of far more than just consumer transportation:, buses,, and delivery trucks will obviate the need for professional drivers and the support industries that surround them.

The Bureau of Labor Statistics lists that 884,000 people are employed in motor vehicles and parts manufacturing, and an additional 3.02 million in the dealer and maintenance network. Truck, bus, delivery, and taxi drivers account for nearly 6 million professional driving jobs. Virtually all of these 10 million jobs will be eliminated within 10-15 years, and this list is by no means exhaustive. But despite the job loss and wholesale destruction of industries, eliminating the needs for car ownership will yield over $1 trillion in additional disposable income – and that is going to usher in an era of unprecedented efficiency, innovation, and job creation. A view of the future Morgan Stanley estimates that a 90% reduction in crashes would save nearly 30,000 lives and prevent 2.12 million injuries annually. Driverless cars do not need to park – vehicles cruising the street looking for parking spots account for an astounding 30% of city traffic, not to mention that eliminating curbside parking adds two extra lanes of capacity to many city streets.

Traffic will become nonexistent, saving each US commuter 38 hours every year – nearly a full work week. As parking lots and garages, car dealerships, and bus stations become obsolete, tens of millions of square feet of available prime real estate will spur explosive metropolitan development. The environmental impact of autonomous cars has the potential to reverse the trend of global warming and drastically reduce our dependence on fossil fuels. Passenger cars, SUVs, pickup trucks, and minivans account for 17.6% of greenhouse gas emissions – a 90% reduction of vehicles in operation would reduce our overall emissions by 15.9%. As most autonomous cars are likely to be electric, we would virtually eliminate the 134 billion of gasoline used each year in the US alone. And while recycling 242 million vehicles will certainly require substantial resources, the surplus of raw materials will decrease the need for mining.

But perhaps most exciting for me are the coming inventions, discoveries, and creation of entire new industries that we cannot yet imagine. I dream of the transportation cloud: near-instantly available, point-to-point travel. Ambulances that arrive to the scene within seconds. A. A merging of city and suburb as commuting becomes fast and painless. Dramatically improved mobility for the disabled.

On-demand rental of nearly anything you can imagine. The end of the DMV!

It is exciting to be alive, isn’t it? It amazes me that the huge, ancillary effects of technological unemployment have generally been ignored. I must admit that I am one of the writers who has done so. Kanter for bringing them into focus. That said, it seems to me that the effects will be even worse than he has stated. Economists have historically estimated the “multiplier effect” of someone being hired or fired as being five times their own income.

(This because people spend their income, which employs others, who in turn spend and thereby employ others.) This suggests that the actual effect of 10 million jobs disappearing could be closer to 50 million jobs lost from reduced PERSONAL spending. And he doesn’t quantify the jobs that will be lost due to DIRECT effects on the various industries he cites (insurance, aftermarket, etc.), which represent many millions of additional jobs that will be lost. Further, I didn’t see mention of US vehicle manufacturers, only aftermarket companies.

Those vehicle makers will be devastated by this, and so will their counterparts overseas. (Of course, this will be a worldwide phenomenon, and Kanter only addresses the US aspect. It’s bad enough even at that.) The disturbing inference I take from all of this is that not only will technological unemployment hit civilization even harder than I and others have been projecting, but that a single profession being wiped out–a profession almost certain to go in the near future, since the technology is already here–could be enough to begin a Great Depression. (I have written a book about a new solution to technological unemployment, and blog about this at.) •. On the upside someone won’t have to spend much to go from point A to point B. Unemployed people as long as they have food, shelter, communication, etc. Can get used to enjoying a nice day at a park with friends and family.

Capitalism will nearly have run its course once green energy takes hold and transportation is automated. Ever increasing efficiency eventually eats away at itself but that will leave us a society where our wants and needs are met inexpensively – music, movies, food, travel, etc. Incomes, stock markets, and home values cannot rise perpetually and that’s ok.

But I’m not looking for a fail safe world where nothing can’t ever happen. Everything is about compromise. But does one ever ask the question if we are compromising too much? How will our experiences suffer when everything is done for us? You may or may not be giving this question its due credit.

Just two years ago my family and I went on a small road trip to Utah. The roads are wide open in many of the areas. No traffic, no nothing. We packed the car as we wanted.

We made a mess as we wanted. We stopped along the the dirty dusty roads for pictures. We drove down near streams and made a picnic.

It didn’t end. The freedom of driving your own car is just that, freedom.

It’s priceless. And just this 4th of July, thousands of people parked along a golf course. Children everywhere we sitting on top of the cars to see the fireworks. Like I said, these small experiences are priceless.

If you really want to protect people, it is almost endless how many of our everyday experiences can be taken away from us. I agree that EXISTING high volume fixed rail mass transit systems (the London Tube, NY Subway etc) will continue to play an important role for years. However, building NEW fixed rail mass transit systems is irrational. Fixed rail systems are inflexible and unable to adapt to changes in demand that changes during the day, week or over the years. High capacity vehicles that are not tethered to the path of a rail line are far more flexible. In the “Network Economy” that we now live in, it is easy to know where every passenger is and where every passenger wants to go at any time.

Transportation assets (ranging from small local pods to high capacity high speed “buses” on dedicated roadways) can be optimized to meet demand. In the interim it will still make sense to feed passengers to the existing mass transit stations, but the inconvenience of fixed rail will become evident quickly.

City planners on the verge of building fixed rail systems would do well to consider more adaptable BRT systems. There are useful ideas in this post. However, it makes a common error. Distribution and adoption rates that would be typical in the rollout of digital goods are fast, and one’s we increasingly witness.

But adoption rates in infrastructure occur over much longer timelines. AV in particular face thorny, and potentially steep regulatory hurdles. The post breezes past, or doesn’t mention, such issues.

The argument would have been better anchored to reality by comparing the proposed rollout of AV with two or three other infrastructure type technologies like wind and solar power, and EV. You know what they say: the first 1% is the hardest.

AV also will face myriad public review hurdles, and insurance problems (The Trolly Problem). I’m very constructive on a future of AV, but one that really doesn’t even get going until after 2025, not before.

Again, examine more closely the long slog that faced wind and solar power, or better yet, the painfully slow sales rates even now of EV. Forecasting AV is hard, but asserting a revolution in this area in a decade is exceedingly aggressive, and unrealistic.

Ever heard of the Jevons Paradox? In economics, the Jevons paradox (/ˈdʒɛvənz/; sometimes Jevons effect) occurs when technological progress increases the efficiency with which a resource is used (reducing the amount necessary for any one use), but the rate of consumption of that resource rises because of increasing demand.[1] The Jevons paradox is perhaps the most widely known paradox in environmental economics.[2] However, governments and environmentalists generally assume that efficiency gains will lower resource consumption, ignoring the possibility of the paradox arising.[3] •. Thanks, Zack. What a great article. I’ll be sharing with a few friends. Well written.

How will all of the autonomous cars on the road mix and mingle with the legacy vehicles on the road? A couple points this article gets me thinking maybe it’s time to buy a car built in the 90’s or earlier. Something not electric but very much manual. Something to be nostalgic about. Will there be more of these kinds of cars? What about Recreational Vehicles or pickup trucks that tow our toy haulers?

Curious on your thoughts on these topics. With gratitude, CG •. Autonomous cars already exist since decades and is called taxi cabs, as well as all means of public transport. Mercedes, Audi, BMW sell every year 2 Millions cars each. This does not respond only to a transport necessity. It would be like thinking an Iphone is just a mobile phone to call your grandmother. Autonomous cars could replace in long term taxi cabs and cover some necessities (I own a car and use from time to time the services of a car sharing company).

And probably autonomous driving would take place in specific roads. I really doubt to see in life (completely) autonomous cars in Naples, Rome, Madrid, Mexico DF, Istambul. What about these scenarios? I’m curious how you envision these commonplace vehicle uses taking place in your version of the future?

1) Running multiple errands with lots of stuff in one trip. Mailing boxes, returning items to the library, returning items to shops, etc. 2) Children in carseats. Children are supposed to be in car seats until they reach a certain height and weight, so it’s not just babies in those carriers with handles. Lots of stuff, frequently both far away and without cell service. Interesting article, to be sure, and I can certainly see that there could be benefits from self-driving cars, but you seem to be assuming a bit too much. To wit: 1) Driverless cars will not need to park.

Unless they’re also riderless cars, I think that might be off. I’m sure they’ll do a better job of it than most human drivers, but they’ll still need to park occasionally, I’m thinking.

2) Autonomous cars will mostly be electric, so that will drastically reduce our dependence on fossil fuels. Cars that run on electricity still require electricity, and the vast majority of U.S. Electricity still comes from fossil fuels. I’m not saying we can’t figure out a way to get most of our electricity from wind, solar, etc.

And figure out a way to store it (and in fact I hope we do) but these are two somewhat separate problems. (Granted, this is not so much an assumption but rather a conflation) 3) The automotive insurance industry will die.

I agree it would die in the sense that it would not look anything like it does now (assuming the governmental regulations which currently govern auto insurance would keep up with the times, which is, to be generous, unlikely), but I don’t think it would go away. Computers break down; not all accidents are the fault of human error, etc. It would likely be much reduced from the behemoth it currently is today, but I’m not sure it would collapse. There are also the broader moral/ethical algorithm dilemmas that have been mentioned (e.g. Should cars be programmed to be utilitarian or to protect the driver/passengers at all costs, to use a basic example); I would be honestly curious to get thoughts on those as well.

Some good points, but some total errors: 1) The title ‘Destroy 10M jobs’ is misleading. A greater number of jobs are created elsewhere. Destroying 10M and creating 20M is what will happen.

2) 99% reduction in cars? That is just far too extreme. I think a 20% reduction in cars is more like (which still eliminates traffic jams, as congestion rises as a square of the number of cars). The good points are that traffic jams will be less, and the vast amount of land devoted to parking lots will be freed up for better uses (a particularly valuable thing for California, where Real Estate is super expensive, yet tons of land is devoted to flat parking lots that are empty most of the time). This story although intriguing and possible on a smaller scale cannot possibly predict accurately with political and economic certainty (even within ten years) much about the future of industry, especially concerning something as personal as car ownership.

At best maybe the industry of driver less cars may be able to survive as a competitor while the business of Uber will expand via leasing cars to individuals and businesses.True Electric/ Solar powered cars will be on the rise even at Toyota and GM, but how about a plan to fix the infrastucture all these vehicles will run on? How about a man on the moon just so we can believe it and watch it in UHD. There are quite a few holes in this article. I will start with the obvious. A car is not just a transportation device. It is a status symbol, even an old clunker is highly prized. Here is my analogy.

In the old days there used to be a thing called a pay phone. It was amazing. They were on every street corner. For a dime or a quarter you could call anyone, order pizza, make a plane reservation, no need to carry around a charger, pay a monthly fee or worry about a poor connection.

Basically all the things one does with a smartphone and for pennies. Yet for some reason everyone would rather pay hundreds of dollars per month for their own portable phone. On top of that most cities have these great things called buses and subways that take you just about any place you need to go. So why have any car-based transportation?

The next issue is, how are people who live outside of town or in rural areas expected to get around? Wait for 30 minutes for the automatic car to come? How about during an emergency when you need to get someone to the hospital? I certainly see automated safety features becoming a normal feature on a car, but few people are going to give up their vehicles.

Just after World War II, the media, particularly the tech media, was filled with stories about the coming age of flying cars. Their argument was based on all the men who’d learned to fly during the war and would be eager to continue to do so. If failed to take into account two critical factors. For most of those WWII pilots, flying had not been an enjoyable experience. They had no desire to continue it. Flying is far more complex than driving a car.

Collisions are far more dangerous, as are engine troubles. The weather matters much more at 1,000 feet than at ground level. Those risks and complexities far out-weighed any advantages of flying, particularly as the country built more freeways. Self-driving cars face a related set of issues.

Which is easier, driving a car or programming it to drive itself? For most of us driving is easy. We get in and go. It’s enormously stupid to pretend otherwise or that we’ll be able to derive much benefit from little five-minute spans when we don’t have to drive but are nevertheless confined to a seat/ For trips around town here’s no advantage to be derived from a self-driving car and there’s the hassle of programming the stops we happen to want to make today.

Only long-distance, freeway driving makes much sense. Long-distance trucker should worry, although I suspect they won’t lose their jobs. Semi-cargos are very valuable, they’ll ride along as security guards if nothing else. How many people who’re gushing about self-driving cars will really trust them? I suspect fewer than today’s polls suggest. Personally, I’d rather drive a car than sit there watching what it’s doing, ready to intervene it it does something stupid, particularly off well-traveled routes. Nor does that belated intervention bode well for driving safety.

Pilots who, in an emergency suddenly have to take over for an overwhelmed auto-pilot often do poorly. To do something well, you need to do it a lot. How much of that 4% utilisation overlaps?

I don’t know how it spreads across the day, but a lot of people want a car for getting to work, or getting children to school. You can’t just think we’ll have 1/25th of our current cars when there isn’t even distribution of use.

Is a travelling salesman going to want to wait around for a car to pick him up between each client visit? With regards to freight, who is going to get a sign-off for the consignment arriving? Is the truck going to get the right palette out for the right customer? I love the idea of self-driving vehicles, but there’s a great deal of unrealistic, impractical things being said about them right now. We aren’t even at 100%, or close to it.

So, you still need a driver behind the wheel who has a license and is awake and sober. At which point, you aren’t making savings from the self-driving car.

Interesting, but very, very speculative. Some of the claims are a bit dubious as well. I imagine there are cities where a significant percentage of traffic is people trying to park but I am very skeptical that is anywhere near the average for all cities, either worldwide or in the U.S. The study cited was two places in ONE city.

I doubt manufacturers would make the safety claims cited here. The product liability exposure would be unimaginably high. Commercial airliners have a much less complicated operational environment than a car in an urban setting and would seem a better candidate for automation (Much of a flight already is automated.). Yet the important decisions, takeoff and landing, are done by humans. I suspect it will be much the same for cars – some backup features for long, monotonous stretches but the important decisions will remain with humans for a much longer period.

As another commenter already said, there are many points of failure as well, not just EMP-sensitive chips, but sensors and code. Even when the technology is there, you need acceptance. How many people would trust standing in a lot full of robot cars going through maneuvers? Then there’s the potential for mayhem by a nefarious hacker operating under the pseudonym “Stephen King”.

Very interesting and thought-provoking post. I apologize in advance if another commenter has mentioned this (no time to read thru all 250+ comments), but there is a fact listed in the post that is incorrect. Uber (and Lyft) drivers do not take home at least 75% of every fare. Nowadays, both services charges fees on top of the regular fare that are not shared with drivers.

For example, Uber has a “Safe Rides” fee. Here’s a recent example in Chicago: Fare for trip is listed as $10 — rider pays $10 to Uber. But how that breaks down -$1.20 for Uber safe rides fee (we won’t get into a discussion here of how much of this is pure profit for Uber) -$.30 tax to the City of Chicago So, the actual fare that is shared with the driver is $8.50. Luckily, in Chicago, Uber only takes 20% (I feel sorry for markets where it is 25%).

-Uber takes $1.70 in commission (20% commission off of the $8.50) -Driver takes home $6.80 of the $10 fare. So, even in a market where the Uber commission is only 20%, the driver takes home 68% of the $10 fare — not at least 75%. Uber pockets 29%. City gets 3%.

Just wanted to update you. Making something cheaper means it will be used more, not less. For example, when the power loom replaced the cottage industry of spinning and weaving, more clothes were made, not less.

Medieval people were lucky to have one cloak. Now, most middle class people have a closet full of clothes. Likewise, autonomous cars will multiply, freed of its most expensive component, the driver. You can expect the streets to be clogged with autonomous cars and the road network to be expanded. For example, before cars, most roads in the US ended at the edge of town. After cars, vast highway networks and farm-to-market roads were built.

Autonomous cars will mean more goods will be delivered cheaply. Delivery services will become far bigger than they are today, with robots picking and packaging the orders, autonomous cars delivering them, and home robots stashing them away. People will subscribe to car services rather than own a car. As cabbies can confirm, there’s also the “peak usage” issue. Even if the claim that people use theirs cars only 4% of the time is valid, for how many people is that 4% of the time, say, between 6-8 a.m. And between 5-7 p.m.?

Also, auto cars will still have to park – the alternative is they drive around aimlessly waiting to be summoned. The idea that they will just go from fare to fare all day ignores the peak usage issue. Further, who exactly will clean them between fares when the guy ahead of me spills his coffee all over the front seat? Also, will my car preferences be similar enough to those of others that my chosen car will be summonable on 30-second notice? If the weather’s nice, can I have an eco-friendly autonomous motorcycle instead?

Can those stupid small cars fit three govt approved child car seats and two adults? What happens when we all want to go home at the same time, 5:00pm? Can I tell my boss I was late because the Uber car was late getting to my house? What about winter weather?

What if I am on call and have to go to work immediately for an emergency? What if I feel like going for a drive?

I hope you like driving at 70 mph on the free way when a plastic bag blows across the road. Anybody who works on machinery and/or computers and/or computer software will be skeptical of tens of thousands of self driving cars driving around the city. People who refer to airplanes that are highly automated ignore the fact that they have two pilots on every plane, “just in case.” It would be fun to be the the only human driving your car with nothing but courteous robotic cars around. Just cut them off. They’ll let you in.

Tell me again who is going to make sure the cars are clean? If I drive a Uber car to a romantic, perhaps illicit, rendezvous, can I be sure of privacy? This sure sounds like a wet dream for a central planner. This will make robbery quite easy.

The perp just steps in front of the car or pushes a cardboard box into the road. BTW, I thought the internet was going to make working from home so common it would reshape the workplace. If that had happened, we wouldn’t need so much mass transit.

The Business Insider infographic stating “90% of these accidents were caused by human error” should not result in the conclusion that there would be 90% fewer accidents. The remaining 10% simply aren’t blamed on the driver.

That doesn’t mean the autonomous car would still crash under the same conditions. In the end, accidents will drop far more than this. Even 1% of the current number would be a high estimate. The rest, time will tell. At some point, the cost of ownership compared to per mile Uber type cost might very well reduce ownership by a high margin. This kind of futurism is familiar and fatally flawed.

It takes a trend and a technological expectation of how it will be used – while ignoring the social factors which are far more relevent in determining what will happen. Allow me to make a different set of predicitions: A) Most people will continue to own personal (autonomous or not) cars. After all, *ride-sharing is very economical today*. If economic logic was the determining factor most cars would already not be on the road. People like having their own personal cars and can afford it – personal cars would always be more available than a shared one regardless of how good the algorithm is, they like their privacy and so on. And the car companies want to sell as much as possible which is maximized by private transport – not shared transport.

Both incentives align, add in the US’s individualistic culture and the result is obvious. [Edit: note also Uber can use an algorithm to replace taxis *today* – this development doesn’t need autonomous cars at all, but it still doesn’t happen]. (Btw, the research quoted in the piece is misanalyzed – the car ownership research is on a self-selected population – people who already use ride-sharing. It ignores that the use of ridesharing is in the single-digits%. The poll is badly phrased: Of course people who use an autonomous car won’t be driving – the car will do it for them!) B) Congestion will *increase* inside cities, remain the same in suburbs and be reduced in inter-city transport. It will be a more pleasent experience though. Humans aren’t the best of drivers, but city transport isn’t congested because of that but because the main transport mode is inefficient.

Since people will continue to own cars (see above), and autonomous cars will allow many more users (elderly people, the blind, even children – why not since they won’t be driving?) and uses will increase too (thinking picking children from school while parent are at work, or sending a zero-occupant vehicle to pick up deliveries), congestion will increase inside cities. Don’t forget people will care less how long it takes to search for parking, so one of the main drives for reducing congestion will be eliminated. Autonomous cars can drive much faster in a congested environment since they’ll negoitate an higher speed – but in cities this won’t gain much (a big margin will still be needed to prevent running over a child or an animal who may just jump into the road). C) Inter-city transport will be revolutionized. Do autonomous cars actually need a speed limit?

I can’t see anything preventing an autonomous car from driving at 200 km/h. Only one thing – electric cars won’t be able to do this (batteries won’t get that good for a good while), and the government may not want to advantage oil-consuming cars that much. [Edit: A few more: D) Autonomous cars will end up much cheaper. The decrease in accidents will lead to a decrease in safety requirements making cars much lighter (this can be marketed as making them more energy effecient).

This will of course make personal car ownership much more likely. Also fuel/electricity consumption will be reduced. E) Autonomous cars will replace manually driven ones somewhat later. It takes about 20-25 years to switch the entire car fleet. It they were available on-mass today the 2035 date would be possible. Hi Zack, I really enjoyed reading your article it was truly detailed and very well structured.although i did predict a different conclusion. My question and concern is, What will happen to the millions which will be unemployed?

Is anyone doing anything? We are being ruled by a few people in this worldwho in fact only think about making money. If their concern is getting richer and the poorer are in fact getting poorer. Then where is the back up plan for all these unemployed people?

These jobs are all jobs which Immigrants and new comers into any country do for a living (bus driver etc), and not only them people who need these jobs. What about us driving loversI would hate to think that ill be alive when this happens and i can no longer enjoy my manual car. I understand the basic’s to avoid running over someone and brakes to avoid accidents. Yes it would be great when I’m stuck in traffic and can only move 30 cm per minute that i have an automatic option which would allow the car to this stop/start option for me. What about robots replacing most jobs by the year 2023? I would really like to hear about your opinion and other readers too.

Thanks, Natalie •. I think the estimation of autonomous vehicles is grossly over rated and that quote from PriceWaterHouse about reducing cars on road by 99% Is a blind dream worthy of an insult for 3 major MAJOR reasons.

Reliable “cloud” (Internet) uptime. Self expression (freedom). Here’s a closer look at those three: 1.

Sprawl – the oil, gas, and automotive industry (the richest leaders/influencers in the world) have been shaping our world for decades upon decades, to become dependent on vehicle ownership, and promoting “sprawl” versus building communities vertically, and for obvious reasons of financial gain, even though they brought their sprawl strategy to life as “the American Dream” to own land and have a big grass yard the kiddos can play in with your own private pool and fence line. Yeah, The more miles apart your commute is, the more money that every single sector in the oil, gas, and auto industry makes! And trust this All those people who live in suburbia absolutely love LOVE owning cars.

Cars of all sizes and all styles and all speeds. There’s no way in hell that it will be feasible to have a variety of 6-passenger mini vans for soccer moms, four door sedans for business execs, two door convertibles for singles on date nights, and all the other variety of needs out there.

FORGET ABOUT IT. We are sprawled and it’s ridiculous to overlook the diversity of needs and the ability to accommodate them in a timely manner with the idea that the right vehicle would be within reach as quickly as owning one. All of suburbia will still have their own vehicles, so that rules out this idea that the Uber concept will have the world careless, which only leaves the question of “do they want to drive uncontrolled (autonomous)?” 2. Reliable Internet connection – we all know about those “dead signal” spots around town. We all know about the cloud and the Internet not always working properly. We alllllll know that even Target and the IRS government cannot prevent being hacked! But we are going to trust our lives and our children’s lives with the Internet and technology at 70 mph?

THERES NOT ONE SINGLE SERVER COMPANY I KNOW that will claim 100% uptime. They all claim 99.99% uptime, and trust yourself on this one, you’ve experienced much more Internet downtime than.01%. I don’t know about you, but I want the ability to yell aty driver when he/she falls asleep at the wheel.

Self expression (freedom) – color, comfort, sound proof, speed, features, options, sunroof, convertible, 5 speed shifting for enjoyment, control!!!! To think that “image” is a thing of the past and that 245 million cars on the road will go down 99%, is GROSSLY misunderstanding the entire spectrum beyond the idea that autonomous means no more accidents. Those 3 MAJOR components need some attention •. I generally agree that when autonomous cars will be available some industries will be distrubted. Some points based on your post. Free Download Saitek X52 Pro Profile Programs For Troubled there. I can’t think why Tesla as a startup automaker will survive compared to e.g Toyota. Just because Elon agrees that autonomous cars will be the future doesn’t mean that the Toyota boss isn’t thinking the same thing 🙂 Furthermore, my thought is that more probably the logistic sector will have “problem” because it’s more “convinient” to replace the lorries/trucks that travel constantly on highways.

Several trucking manufacturers already introduced lorries that can accomplish the task. Perhaps this is the sector where will we actually witness the transision to autonomous cars?

“The effects of the autonomous car movement will be staggering. PricewaterhouseCoopers predicts that the number of vehicles on the road will be reduced by 99%, estimating that the fleet will fall from 245 million to just 2.4 million vehicles.” I think there is an error in the source.

It talks about 90%, and then has a chart where it takes 90% of Accidents and Energy, but then uses 99% to go from 245 million to 2.5 million cars. If as another statistic sugests we use the current car base at a 4% rate, and we move that to a perfect 100% rate that would drop us to 10 million cars. I don’t know what a good estimate for usage rate on a fully autonomous system would be but I would be supervised it it hits above 50%. Also if transportation drops in price, increases in convenience then I would expect a greater utilization. I do agree with the rest of the potential. The next question is what creative political/legal blocks will be created to slow this process down.

Great article Zack! I really enjoy seeing how we keep moving towards a next vision of auto-making, transport-thinking, personal-to-collective but without missing the point that people (or a big part of them) will always have a need for distinguishment. The “showy” side let’s call it. We focus on collective in order to reduce mass.

Electric in order to reduce pollution. Self driven in order to reduce accidents and safe quality time. It’s a fact that we are going electric and there is no doubt of it, and going collective will increase although many will still want to “show”, to “position” themselves on a personal car.Looking at it from another side, and if a major goal for transportation will be to reduce lives, what will happen when we will have reached the goal (zero death on a car)? This inflection point will force us to go from reducing injuries (which could translate into saving lives), towards “generating live”. Why not then having personal cars/cells which keep supplying (solar electricity, prifying air) while not being used. In my opinion there has been a tendency of thinking how a car could do better for its last recipient: people.

I like taking the last recipient as “the Earth”, “the City”, and I never heard about seing it a “personal machine” which “while” generating, producing, supplying, etc, can take people from A to B and make them feel unique by “giving” rather than by “showing” (a way of show-off turned up side down). Great article. Well articulated vision of our inevitable future. I wonder though if it oversimplifies the abondment of car ownership. Will parents forgo their own trusted child seats overnight?

Probably not. What about people who need vans or pickup trucks? I don’t know the figures for how many cars that adds up to but my guess there are many such examples that aren’t being accounted for in all these futuristic predictions. Like anything else, I believe there will be a balance between the old and new world. The automobile industry will not collapse, it will simply have to adapt.

Perhaps people will pay a premium to have a professional driver in their Uber to converse with. Who knows We’re not that predictable 🙂 •.

Your article gets picked up by QZ, wired and Inc. And that worries me. Firstly, because these wide reaching publications are spreading simplistic narratives about a mechanistic future that does not take into account human nature or the way economies, societies and policies affect changes. Technological meme – teme are powerful, no doubt. But they can be nudged.

How do you think the megatrends about concerns for privacy and control will play out? How would the labour (drivers) would react? In democracies where they vote, what policies would affect adoption of this tech?

Is 90% self-drive car, good enough? There is a problem with current crop of ‘visionaries’. Their starting point for intellectual exploration is ‘what can technology do?’ instead of ‘what should technology do?’ Technology needs a human centric approach. With autonomous cars, the tech giants are answering wrong question. Tech should right now be employed in making commutes redundant.

We don’t need driving less world – we need a car-less world. We don’t need disruptions. We need evolutions. Here’s my perspective. I will give you an example. ‘what will a hammer do?’ That question has no meaning. Hammer has no agency.

Hammer is a technology. Autonomous car is a technology too. It ‘will’ not do anything. Individuals, societies, companies and other such organs of human societies ‘will’ do something with the technology.

Future gazing devoid of sociological, economical analysis is a waste of time. Here’s another train of thought. If car is extension of our limbs (McLuhan) Then an autonomous car is an alien limb? Who will want such an extension. Here’s my prediction. Autonomous car will be the segway of 2025. Playthings of rich people.

The inequality isn’t going to reduce in next 10 years. This is a technology that only worsens it. Autonomous cars will only remain status symbols.

Unless, there is a major revolution in town planning, immigration reforms and basic tech behind autonomous cars, i don’t see a future of it. Right now, the tech behind autonomous cars is very nascent. To know where it should go, it employs far too many resources. There are way too many variable not yet resolved. Unless the tech becomes smarter, it will become the Delorean of 10 years from now. Developments are going so rapidly, that you start wondering if automakers are really paying attention. Believe it or not, to them it is still about market shares and how to squeeze more out of modular platforms and technology, than to come up with a new let’s call it “automotive format”.

Even if 50% of what Zack says is true, then the challenges are actually pretty much defined. Needn’t be the end of the car as we know it. To give an example, Apple reinvented the good ol’ phone that much, that ownership was taken to a whole new level •.

“the $300 billion automotive aftermarket21 will collapse as demand for their services evaporates.” what aftermarket do you refer to? I was in aftermarket business many years ago, virtually every customer “wanted” to spend money on their vehicle just for the sake of being unique. Just because a person has the disposable income to buy, what will be a very expensive, autonomous vehicle in 2020 doesn’t mean he won’t spend money to make it look different or move faster than the next guy.

Other than that point I enjoyed the read. I think there may be both over and under estimation going on here. After all, where are the flying cars we were promised we’d be using by now?

There are what, 200,000,000 cars in the US? Think they’ll all just vanish in the span of five years or ten or fifteen?

I especially don’t think so in the airier (less dense) regions of the country. Maine, Vermont, Wyoming, the Dakotas, rural Texas. It’s not possible for the bulk of the country to be 36 seconds away from an empty fifty-cents-a-mile cab. And for those of us driving 10-12-15-20K miles a year, it just doesn’t make sense to be hoping that we’ll be able to get a cab any time we want one that can carry us and all our stuff where we want to do. We don’t want to go uptown- we want to go to Cumberland, and then stop at the library in Freeport and then head to Bath to pick up a kid from his after school program.

The 4% thing is irrelevant. How often do you use your umbrella?

How about your furnace? As for $9K a year We run six cars (wife, self, four kids, no public transportation, nothing close to anything.) and the insurance is a couple grand, the repairs two or three, and the cars are all more than 20 years old.

It might cost us $1K-$2K per car for those that get the most use. So maybe $9K max, total, to run six cars. Cost of living in a small town in a big, rural state. The flying car thing is a bit of red herring. But I agree about the installed base of cars. When Zack says autonomous vehicles will “have a near monopoly by 2030,” I assume he means a near monopoly in new vehicles sales.

Because just a little math shows that a prediction of even half the fleet being autonomous by 2030 requires incredibly optimistic assumptions. Cars last a very long time these days. My car is 20 years old and I’ll be surprise if I don’t still own it in 2020. (Unless I simply don’t own a road car at all by then, which is possible.) •.

Used to be 7-8 years was the norm. My newest car is 22 years old, my oldest is 27.

I concede I’m not representative, but yes, Cars are lasting much longer, and I think there will be more resistance to autonomous cars than some. I think in places like California, Chicago, New York (City), Atlanta, Houston– you might see a lot. But out here in small town land? I know one shop that specializes and the cars they currently sell went out of production in 2008. He sells 4-7/week, no shortage of buyers.

There are not yet as of my chime hundreds of comments about your autonomous vehicle entry here but there are too many for me to yet claim I have parsed fully all- the loss of income you are suffering giving away this text to nextgov etc. To run adds on is obscene ironically however much so though son. For me, and I do keep up with technology and competitive intelligence to use an old if not archaic phrase, the excitement which both versions of your writing above end on needs to be moderated with historical appreciation of the work we must do to have the potential good be realized. We have parking garages, and incredible amounts of the earth paved presently.

The percentage of our earth devoted to cars actually increases as open space is found lacking- a true riddle perhaps the greatest one in it’s lack of recognition. Those who push for higher densities mean in fact more roads, less people, in the sun.

Speed causes us to be deluded about having space to drive. Having space means you can drive slow.

Few of us are allowed to drive any distance slow. Having space also allows great distances to be covered in short time. Although the need to do that more then every few years at most has been vanishing for many decades.

For decades we have been denied the revolutionary disruptiveness of the street car. Without a free internet GM was able to so pollute public discourse as to still deny us them so many decades after they got convicted criminally of exactly that. Yes machines can drive better then us- in the last century or so that includes steering and for decades as well navigating. Yet one needs to goto Singapore nearly to flush a toilet without using drinking water, or to poop with any rationality nearly into space. For me leaving the atmosphere should have little to do with being able to use decent toilet. So it is with cars. To make them fly as expected, is mundane.

To be able to teleport just virtually lame. To be able to access what smart educated people can figure out to make things better for all of us? That is where we have not recovered yet from the first libraries getting burned down. To be very clear to concern yourself with such a topic is to be a crack whore to the military industrial etc.

As corporations attempt to intercede where Uncle Sam will allow now our best and brightest to do good works again but for them, we must strike. See todays WSJ about how they want to further plug the gates of Harvard with there agenda, not the public’s. We used to be a society of single wage earning households- and we can become one where such are the minority and to be a good parent is to be one where your children will never have to sell themselves in anyway. In such a world nobody has a large staff devoted to flying them around for fun. Not even movie actors. There is no starvation, war, car lots with high-school drop outs waiting to attack teens and little old ladies with easy terms.

Rather Rolls Royce’s rule- they are for the poor to sit in to get someplace for some reason they can not afford to walk to. Economic mobility is best had by mobility by need only being realized. Napalm in your tank, coal plumbed by any means, is to eat dirt, and grovel for survival. Such is GM’s present propaganda, so is our Tienanmen square being set to slaughter yet again.

The rest of the story is missing, and must be told. What would happen if the USA had the voter registration of Australia, without coercion?

If the camera’s and other ‘enabling’ (think cancer) technology of ‘undriven’ cars instead watched childrens faces to help them better learn? Such choices can be ours. Or we can just blog on. I don’t think so. In rural states, auto usage is quite a bit higher. 4% would be around 7 hours a week.

Without vacation or other long distance travel factored in we drive more than 10 hours a week. Add travel and you’re looking at enough to amount to another 2-4 hours a week.

Then there’s the wait time. No way are you talking 36 seconds. That little bug in the photo? We need space.

All but one of our cars are station wagons. Oh, and I don’t like autonomous machinery. Those station wagons all have manual transmissions. A wonderfully Utopian take on the situation. Though I find it hard to believe that completely autonomous vehicles will even be allowed to operate without a driver, let alone dominate the road.

Aside from government regulation and public acceptance, sadly, every industry that you mention will be put out of business are just a few of the many reasons this will not happen in my lifetime. Any ONE of those industries is powerful enough to stop it, but unite auto and oil industries, taxi unions, truck driver unions, mechanics unions (just to mention a few), and millions of drivers = an impenetrable wall of lobbyists, corruption, and our government at its finest.

Not to mention the tens of millions of people not actually located in a metropolitan area at all. I could see this having a significant impact in MAJOR metropolitan areas (and I mean NY City, San Francisco, etc) within a few decades, but I doubt most Americans are going to see the need. I work in the Sacramento area () with 2.2 million people and yet there is almost no need for this even in downtown Sacramento. All the congestion I see every day are commuters coming from all over the place. UBER is not going to have enough cars for the 70-80% of people (even in metropolitan areas) that commute alone (). And even if they did you are going to have to change the majority of the public’s behavior.

And outside of commute time in most places there is really no issue. What I could possibly see happening is everyone driving (owning?) autonomous cars however. I could sure use a good nap on the way to work!

That would still save lives, fuel, and cut down on commuting/congestion (more efficient driving hopefully). Unfortunately, driverless cars are not the only disruptive technology that is replacing human jobs with computers. It’s happening in many fields: paralegals, medicine, retail, and all the pointers are to increased job losses.

A recent study predicted that in Germany, 60% of jobs will be lost to automation in a few decades. And of course, automation will be making inroads into new job categories as software and hardware get more adept. I don’t believe the magical “job creation” fairies will be saving the job market any more. It’s Basic Income is going to become a necessity. This will never happen. I love the thought of taking a nap on a cross country trip, but the first accident in an autonomous vehicle will be the last. Who is liable?

The napping driver who bought the car? The company who sold the car?

The company outsourced who wrote the software that drives the car? It’s a legal nightmare.

That’s if we don’t ship each one with a bug – a rolling ‘blue screen of death’. Let’s assume we solve all the technical complexities and give legal indemnity to the driver/seller/manufacturer – what about the ‘have not’ crowd around it? The first time someone figures out that jumping in front of the funny looking car isn’t a death sentence we will have a rash of incidents ranging from adolescent prank gridlock to the easiest (and most lucrative) carjackings ever. I want your vision to be the future – but it’s just not likely to come true 🙁 •. Sorry, I dont see many environmental benefits. While the number of cars built and sold will doubtless plummet, the number of kilometres driven will not. Distance travelled is the environmental driver here, not the number of cars doing the traveling!

People will move to these services from public transport, so we’ll end up with more people on the roads, in far fewer cars, which are constantly in motion rather than rusting in their owners’ driveways. Autonomous cars will drive more efficiently, but any net environmental benefits will only appear if these cars are made much more efficient to compensate for the increased usage caused by people abandoning public transport. As there are fewer cars in number – and their running costs will become a competitive element between the mobility operating companies – this should be possible but it’s not guaranteed. An interesting and thoughtful article, Zack. This scenario reminds me a bit of James Halperin’s famous novels where he forecasts a single technology change having widespread consequences and making life MUCH better for almost everybody. Some of this seems likely, and it’s nice to see thinking out of the box.

When speculating about the future where MANY of the specifics are unknown, naturally there will be room for different considered opinions. One thing I noticed was that some of the articles cited (like item 17, the PWC prediction of a 99% fleet reduction — where such grand predictions are not sourced nor explained) make little sense, given other assumptions in the article, or just basic arithmetic. If cars are used 4% of the time, how could the their transportation ability possibly be replaced with 1% of the fleet? 30% less city driving due to less parking doesn’t get you there (not even close).

Besides, such cars driven continuously will need frequent maintenance. (Unmaintained and therefore unreliable cars breaking down will NOT encourage people to give up their private car).

So are we replacing the public transportation network (which moves a LOT of people fairly quickly) in big cities, or reducing the car fleet by 99%? You certainly can’t do both. Also, given the congestion on the roads in cities at rush hour, I have serious doubts such cars would even FIT on the roads. If you can convince (or mandate) businesses to spread their employee start/stop times way out, that will help. However, given the demands for competition and low cost in business, this will have its limits. While we’re talking about replacing the public mass transit network, I noticed the point on ride sharing.

Having to share your ride with random unpredictable people who may be unpleasant or even dangerous is a big negative to the idea of giving up car ownership. If people can’t control this by, for example, choosing to pay more and not share a ride (and perhaps agreeing to avoid peak traffic times to secure this right) — I suspect this idea won’t be widely accepted unless it is mandated. For people who are frugal and don’t drive a lot and buy a low cost efficient car and keep it a long time (i.e. Take care of it), the annual cost of having a vehicle can be much more like HALF of the $9,000 average figure you cited. (I’ve calculated this for myself, and being retired, I drive about 6000 miles a year, 99% of it locally).

At your optimistic estimate of 50 cents a mile to pay for an Uber-like service, that $4500 figure will only take you 9000 miles, which is less than the average person drives today. So for a relatively modest sum, financially stable people could choose to keep their own cars, unless government mandates otherwise, or sets up strong financial disincentives (like extra taxes).

I’m not at all convinced the auto insurance industry will be “dismantled” unless your vision of almost no private cars comes true (and given human nature, as I said in number 3, I have serious doubts). Changed, yes. Cheaper — yes, if autonomous cars are meaningfully safer.

However, that’s far from “gone”. So again, nice job. It’s nice to see the positive, polite tone of the comments I’ve read thus far. I think the 99% reduction is just a typo in the cited source. The article meant to look at a 90% (not 99%) reduction as a assumption, not as a conclusion. I do think it is realistic to consider that some point only half the cars we have now will be required, and that is still a massive drop in all the associated industries. Also this will not be equally distributed, and those places that can drop the fastest will gain an economic advantage over those place where it drops less.

If shared cars become the norm, then i expect it will open up the range of vehicle design to tailor cars to a tighter range of needs at different price points. For instance a low cost shared car that caries multiple unrelated people over longer distances could provide different compartments so those people can have some privacy on the trip. Or it could be that lots of smaller 1 person cars would be more efficient. I can imagine cars designed to go two abreast in the current road like a motorcycle being a good low cost solution. “But despite the job loss and wholesale destruction of industries, eliminating the needs for car ownership will yield over $1 trillion in additional disposable income – and that is going to usher in an era of unprecedented efficiency, innovation, and job creation.” You seem wildly enthusiastic about this Brave New World, but have you thought this through? Two things do not seem to have occurred to you. First, autonomous cars will obviously be the delivery system of choice of terrorists.

Second, the widespread adoption of autonomous cars will inevitably entail further erosion of personal freedom. Perhaps that does not bother you. It bothers me. So terrorists can’t take the bus or walk or drive their own car or take a taxi? Somehow I don’t think the fact that there are some bad people in the world will make or break this technology. Should freedom to stand in traffic jams and be injured or killed by the many careless or negligent drivers be more important than the freedom to turn driving into a much more efficient and cost effective and safe process for almost everybody? Given the trade-offs and how much other technologies and our government (the IRS, the NSA, the FBI, etc) are ALREADY impacting our freedom — sorry, but to me you are over-reacting.

If people still crave driving, there will probably be things like private tracks where you can go and risk your well being to manually drive a car, at high speeds even. Society is all about trade-offs geared to balance the rights and needs of everyone.

If you’re unwilling to make those, you’ll need to become a hermit. You seem to be intelligent (if perhaps rather arrogant) but your first sentence makes me wonder: An autonomous car – and even more so an autonomous bus or lorry – can obviously deliver a much bigger bang than a pedestrian or a bus or taxi passenger. Thanks to improvements in the design of vehicles in both passive and active safety and also some improvements to ‘street furniture’, road casualties have dropped dramatically over the past 30 years. But I agree that there is still the problem of driver behaviourand also pedestrian and cyclist behaviour (personal responsibility applies to all). It was stupid, for example, for governments to allow the use of telephones (even, or perhaps especially, the hands-free type) in cars. Rather late, they seem to have understood their own error.

I suspect that you are a city dweller who does not need to drive every day and that when or if you need a car, you hire it. But there are millions of people in every country living in rural areas who need cars for their everyday lives. They will not be able to afford to live in your Brave New World. There will be a lot of ‘hermits’ in it, willing or otherwise. It’s an interesting post, but as someone else noted, driverless cars are another aspect of mass transit, not a replacement. Many people who live in cities get rid of their car. I never owned a car when I lived in a city and wouldn’t own one if I moved back.

However, safety is an issue we can’t answer for these cars, yet. What about icy conditions? The rider is powerless to do anything if/when something goes wrong, but that’s also true whenever you’re not in charge of the means of transportation from cabs to trains.

The issue of computer code being wrong is a very important part of this because no code is perfect. I’m sure whatever happens, it won’t be this smooth.

I can see the future clearly now. Self driving cars are inevitable, the car form factor is not imagine if you will a self-driving Segway! You basically sit atop it at a small desk, where you can sit and use your laptop or whatever in a sealed pod that sits atop what we know as a Segway today with a human removed from the controls it can reach 70-80 MPH easily. It’s super compact so it handles even more passenger density than a bus, controlled by a computer you could easily have six abreast in a lane. If the traffic got too bad they could switch to using sidewalks or other pedestrian walkways as viable alternate paths, keeping to walking speed to mix with the few people still walking.

Indeed being contained is there any reason why they could not also use rivers as viable causeways? Good thoughts.

I run a connected transportation operation in Phoenix called RubyRide – kind of a true SaaS for logistics. Uber has a lot of money and name recognition, but it is really premature to assume that they will end up in the lead with the self-driving car, even with the money they’re throwing at it. They have loyal users in some communities, like SF, but in most cities they are purely a marketing group, and as soon as they turn on their driver ‘partners’ they’ll lose a lot of good will. Running a real transportation company requires real infrastructure, operations, etc – things Uber has no experience with. In any case there is a huge opportunity to explore the added value that a true CaaS (Car-as-a-Service) can provide – something Uber hasn’t even begun to look at, and their business model makes difficult.

The technology, of course, is well ahead of the regulatory and market forces. I think it will be more difficult to roll this out than you think. People like their cars, and there will be a long period (10 years +) in which existing rolling stock will rule the road – the vast majority of cars on the road are owned free-and-clear, so the incentive to give those up for ‘rent’ or pay-per-use isn’t so great. I’d like to think that your prediction about a permanent increase in people’s disposable income is on target, but history has shown again and again that the economy is elastic.

When one piece changes, market forces will adjust. Pay will stagnate or rents will increase worker productivity is so much higher today than 30 years ago – everyone thought that we’d have more leisure and less stress •. No technology is perfect. OTOH, things like Toyota’s car making robots have multiple redundant fail safe systems. When you consider how UNSAFE human drivers frequently are, through carelessness, inattention, mistake, or just stupidity or bad judgement — it’s not hard to imagine a point where the driverless car is safer.

And it’s not like there aren’t lots of statistics that can and will be used to measure and compare the driverless vs. Human driven cars. By the way, there is encryption, etc. The car hacking problem you cite is already being worked on.

You can Google it. And yes, there is such a thing as a fail safe: •.

The 99% reduction in vehicle fleet size is probably a typo carried over from ref 17. If vehicles are used 4% of the time now we can’t expect them to be used 400% of the time in the future, I would say.

40% seems a more likely (but still very aggressive) goal. On the environmental improvement this blogger is totally off the mark as he thinks that emissions will go down 90% by having less vehicles, forgetting that these vehicles are going to do as many miles as the original fleet did, presumably consuming as much fuel. Indeed, if these automatic taxis spend their off-hours circulating instead of the current system where cars ar eparked waiting for their owners, it seems likely that consumption will instead go up.

In addition this operation mode where vehicles are constantly in use goes totally against the idea of electric cars which need long off-hours to get charged. Batteries that last a full day of automated taxi operation are still not even visible on the horizon. I have worked quite a lot with simulation of PRT systems which use a similar operation mode on dedicated infrastructure, and it turns out that 20-30% of trips are empty relocations. This number should be similar for automated cars (maybe higher as there are no fixed station locations, so every ride requires a car driving up to where the passenger is). So be a hermit, live in a remote area, and drive your motorcycle. It will be a LONG time before laws prevent that, if ever. Meanwhile, there is a growing net migration of people to cities or jobs and for convenience.

By the way, if you’re afraid of technology, there is a lot of technology in the vehicles, and even the gun you cite. Funny how you’re not afraid of that, since it’s already “normal” to you. Well, in a couple of generations, autonomous cars will be “normal” and most youngters will be glad they don’t have to wait to get a driver’s license to go where they want. Hi Zack, interesting article. I agree self-driving cars will cause significant disruption and many benefits. However I think the key point that you need to look at is that road space is not an infinite resource, and hence driverless cars cannot replace mass transit.

An example may help. I live in London, where our various trains move over a million people into the city centre each day.

Each tube train carries in excess of a thousand people, and on some lines there are up to 34 trains per hour. The roads are already one big traffic jam of 118,000 morning peak passengers going to work on busses. In between the buses we manage to squeeze in a few trucks doing deliveries and one or two private cars. Clearly shifting those million train journeys onto an already jam-packed road network is not feasible, so public transport has to stay. It’s a similar story I’m sure in any large city, like New York or Tokyo. In my view self driving technology will be applied to intra-city mass transit to make it more efficient rather than disrupting it.

Now onto the point of journey pricing. In numerous studies we have seen that supply of transport generates demand. So in smaller cities (maybe like your own) where the roads are more sparsely used, a fare of $0.50/mile will generate massively increased demand, and people will make far more journeys. This will lead to ever-increasing congestion. The congestion will lead to one of two outcomes. The first is pricing will artificially be held at $0.50, and congestion will continue to grow until it becomes impossible to move around.

The other option is that prices would rise to reduce demand (either by government taxation or car owners increasing prices for higher profits). I would guess that prices (and levels of road use) would naturally gravitate to near their current levels. We already have the tools to deal with lack of parking space, road congestion and high levels of car ownership. If we start charging cars for the space they occupy while parked and a per kilometre/mile fee to use roads, only truly valuable journeys will take place. We don’t need new self-driving technology to solve these issues, rather political and popular will. Hey Dean: You make a good point about city congestion & mass transit.

Its the only sustainable answer for crowded cities. The question is: can we come to think of a network of autonomous cars as a new form of mass transit? The network — “transportation in the cloud” — will feature a combination of algorithmic dispatch and pooling multiple passengers per vehicle. (Think UberPool) The dispatch algorithm can minimize 1 person per autonomous car. Or raise the rates on single passengers. City regulations can mandate 3+ per vehicle — and it will be up to the network providers to solve that for everyone. In the end, its a shared service that scales access to transportation and creates efficiency over individual ownership – Mass Transit.

Paul Kogan, CEO, Hovee.com •. So that’s a retraction on this, Zack? ” Such convenience and low cost will make car ownership inconceivable, and autonomous, on-demand taxis – the ‘transportation cloud’ – will quickly become dominant form of transportation – displacing far more than just car ownership, it will take the majority of users away from public transportation as well” Also I can see a bunch of reasons why people would pay more to own what are today higher-end branded cars, self drive, manual and hybrid. Same reasons they spend so much money on luxe brand cars today.

Conveyance of travel is but one motive in the car purchase today and you’ve considered all the other reasons to be inconsequential in your brave new world of self-drive road vehicles. This line of thinking is what I call take-an-emergent-trend-and-double-it futurism.

Is it really worthy of the title futurism to say ‘hey this new invention is going to disrupt the shit out of everything cause it’s just got to’. A lot of people who are paid to think about the future were saying that about Segway before it launched. It’s not that convincing to me and I’ve spoken to a few people involved in modelling public transport systems for the next few decades about this kind of issue i.e. Car vs public transportation with a view to getting a beyond zero emissions economy. Who will pay huge money for a car, when no one thinks its a status symbol any more and everyone laughs at you for wasting money when they can send a text, have a car in 5 minutes, and be anywhere, then abandon their car to do anything they want, and get a new one at any time. Never worrying about parking, safety of the car when left alone, damage, maintenance, fuel, etc there was a time when cars were coming that peopkle just like you said nearly the same thing about horses.

I dont see any horses around. But that’s just it. With less people owning cars it will be a HUGE Status symbol. We will have Autonomous luxury cars. The well off and wealthy will still have cars.

And the hot women will still be riding in them with them. That will not change. Who the hell wants to use public transportation. And then you have people that actually like to drive like me. I’m not going to take a Fing uber car because I like to drive and if I don’t want to drive I’ll have my car drive for me.

But I still want my car so I can take my stuff with me. It makes a big difference which city you’re talking about.

Most European cities already have a comprehensive train infrastructure, and that has its own advantages and efficiencies that won’t be overcome quickly or easily by another technology. But most North American cities have poor mass transit infrastructure, and even the best probably fall short of the European average on an apples-to-apples comparison. Autonomous vehicles will allow North American urban transportation to become much more efficient without having to build mass transit infrastructure. That’s a good thing because those cities are already configured on a road-based infrastructure model that is not easily converted to a mass-transit model. The average person or household is not going to be the driving force (no pun intended) behind the autonomous car becoming more commonplace in traffic.

The first obvious markets are car pooling services, taxi’s, delivery services and at some point even long distance truck deliveries. To put this into a practical perspective. The first taxi company that starts using autonomous cars will effectively remove the majority of staff costs by not needing drivers and not needing the administrative staff to follow up on wages, welfare and other associated costs for hiring staff. Instead they will require a skeleton crew for monitoring the vehicles as well as service and maintenance.

The reduction in expenses means they can also reduce the cost of the fare, thus becoming a more attractive company for the average person, which in turn means any taxi company without autonomous cars will, over time, either have to switch to autonomous cars themselves, or vanish as they are no longer competitive on prices. The same principle will apply for delivery services. No need to hire staff to drive the vehicles means less costs, which means they can lower prices and still earn the same profit, thus running any other delivery service out of business or forcing them to switch over to autonomous cars as well. Not to mention, in both the case of taxi’s and delivery service you can have the cars running, literally, 24/7 without the need for a driver to rest, having lunch breaks et.c.

That means there is effectively no non-productive time for the cars other than stopping for fuel and/or service and maintenance. For long distance trucking this has a double effect since there are restrictions on how many hours the drivers are allowed to drive without resting. For very long runs this is mitigated by having 2 drivers where one rests while other drives. However, with autonomous control there is no need for any of the drivers, plus there will be no requirement in terms of how many hours you are allowed to drive.

Thus the long distance trucks can also run 24/7 and only requiring stops for fuel and/or maintenance. Again, this advantage will easily give the first company to employ the autonomous cars a distinct market advantage. The only thing that may delay this is the cost for the autonomous cars, the costs of maintenance and of course, how well we facilitate for these types of cars to refuel as well as drop off any deliveries. But again, I am personally convinced that once autonomous cars are seen on the streets we will also see gas stations adapt by having pumps that allow for automatic filling of the vehicles as well as pickup and delivery areas either having automated facilities for loading/unloading, or simply have staff to unload the vehicle upon arrival. Will this happen overnight?

Of course not, but over time this will be more and more commonplace. Once the different markets such as taxis, car pooling, deliveries, long distance trucking and so forth raise the demand for such cars it will also lead to more money flowing into the autonomous car industry, which in turn leads to development into making them cheaper, more reliably and thus making them affordable for the average household which means even more autonomous cars in the streets.

First person to patent an automatic refueling pump that works for a large variety of different autonomous cars will be making some money though! Interesting article (if a bit dreamy on the positive effects of driverless cars). I too have thought about the issue of driverless vehicles, and I believe in a longer time frame. Sure, Google and Tesla have been working on bringing the driverless experience to the regular vehicle, but I believe “the driverless revolution” will occur first with Long-Haul Tractor-trailers. Imagine a big rig driving during the ten hours that normally would be spent sitting at a rest stop or truck stop, and the trucker starting his sleep just past Denver and waking up as he comes up to the Kansas City metroplex. Next will be public bus service – Megabus comes to mind, then a couple of BRT routes will either be retrofitted or created with driverless busses in mind, and finally some (southern) city will just say “screw the drivers” and have full-city driverless busses (others will follow). Finally, we’ll see more of the ride-share type arrangements you’re talking about, with people off-loading their ownership to some Uber-like entity for a monthly “rental” fee plus per-ride fees.

Given this line of roll-out, fifteen-twenty years is logical – especially once the taxi companies figure out that a vehicle with a $20,000 guidance system beats having to pay two drivers $30,000+ for the privilege of abusing their vehicle (by driving madly, using it as a drive-around diner, etc.). And things get better.

Consider robotics. At Toyota, they have MULTIPLE fail-safe systems for the robots on the assembly line. This protects workers from being injured whether one monitoring device fails, the stupid human disobeys the rules and signs, and blunders into the robots’ space, etc. If robots were seriously injuring people except in very rare/freak instances, the lawsuits would make them financially impractical. Well, a fully autonomous car will be a robot-car. Same principle.

By the time they would be ready to replace drivers, there will be a LOT of data to show that they are safer, or not. The public won’t have to guess. Given what I see drivers do every week, I’m betting that in time (not right away, necessarily), they’re FAR safer than human drivers. Expect RV culture to expand and change. We will likely see entire communities, connected by shared interests (religion, arts & crafts, surfing, music, etc.), that live in autonomous vehicles year-round.

Individuals and families will sleep in their vehicles and the communities will share trucks with larger facilities for kitchens, bathrooms, recreation, and storage. Many of these groups will migrate with weather and the seasons. I expect some of them to develop bartering economies and set up bazaars for outsiders, but can imagine other groups that can work remotely or on-site as contractors (software engineering collectives, etc.). I’m also anticipating the first highway-based hotel. Let your hotel “room” drive you to your destination while you sleep! Safe attended garages at either end to finish your night’s sleep and get breakfast, and well-maintained, high quality drive-through rest stops along the route.

Incredible article. I started using uber when i made the decision to sell my car, buy a bike in a not-so-bike-friendly city to expand and explore a different sense of freedom. My experiences made me realize and appreciate the offerings of both human connectedness and easy reciprocation of service and gratitude that Uber grants us. I have sensed the major impact on transportation, our lives and ecomony but have lacked the facts and knowledge to bring it all together as you have. Bravo, Im so glad to have found your blog! • Pingback: () •.

You tech people seem to forget something. As robots put us all out of work, where will be taking these cars too. America is going to quickly adjust its attitude that everyone must contribute to society by having a job, or there will be a tremendous amount of turmoil and bloodshed in this country’s future. I for one will not sit on the sidelines while the rich hoard theirmoney and have robots perform most of the necessary work. Expand the “welfare state” and allow us to live our lives freely or this robot nation will have serious repercussions. Amusing read.

However, I disagree with you about public transportation. I see you already concede in a comment that self-driving vans and buses will not disappear. What will Greyhound do?

Currently, one can travel 1000 km by bus for 80 US$. I have a hard time seeing how an individual private self-driving car could go down to 8¢/km (12¢/litre). So, bus stations will still exist.

And your calculations are mainly based on urban driving. Car dependency is fundamentally worst in rural areas, though. People who go camping in a remote corner of Utah will want to keep a car with them for days, they don’t want the car to leave them as soon as they unload their gear, because it might take an hour for the nearest available vehicle to reach them — every time we need it. Therefore, I think car rental is still going to exist. (And I’m told some people actually *like* driving, in particular 4-WD. Already now I know people who rather drive a manual transmission car than an automatic one.

Do you think they will all change their mind?) •. In respect to your last paragraph, it’s more a generation gap issue. Yes, a lot of people like to drive, but their children and grandchildren who’ll grow up in a driverless era won’t even have a chance to acquire the taste, or if given the chance – won’t see the need. This trend is already quite prominent in European cities with excellent public transport networks, e.g.

Vienna; many young people there don’t know how to drive and don’t want to learn how, even though they’d probably like it as an activity overall. But given the cost of ownership and all the hassle of city driving, young people simply take the tram or subway, or ride a bike. Also, as the benefits of driverless technologies on human health (in particular the rapid decline in loss of life) start to show up in statistics, I wouldn’t be surprised if manual driving eventually becomes outright banned by law (with legal wording penned by those children and grandchildren, mind you), or allowed only in very limited circumstances. If you argue that you can’t outlaw something that a lot of people like, please don’t even go there, just give it a second thought.

The majority of people live within a major metropolitan area – but yes, I agree that rural areas will likely hold out for quite a while. I think that autonomous bus transport will be popular between major destinations (e.g. My point about bus systems being obsolete was in reference to intra-city transport – I don’t think that their volume will sustain once cheap, direct, point-to-point transport is an option. And yes, 4WD/racing/etc will continue – just like horseback riding continues today despite horses not being a viable form of transportationjust in far smaller numbers. “a 90% reduction of vehicles in operation would reduce our overall emissions by 15.9%” – surely not, since the emissions come about when the cars are in motion. Whether they are autonomous or not doesn’t matter.

The benefit does exist with less congestion and eliminating those 30% of “parking traffic” (which, to be sure would only apply to densely populated areas, or are we saying that 30% of the time we spend in cars is to look for parking!?). Other than that, agree with the spirit of your post – exciting times ahead! Along with electric vehicles, there’s also the possibility of matching vehicle size to trip-need.

Many many people own very large vehicles, but only use the capacity a small percentage of the time. Even something like a driverless Prius is carrying 4-5 times the passenger inventory that is most commonly needed (you don’t need 5 seats for 1 person). If safety rates increase like they’re predicted to, it’s not inconceivable that something the size of an electric driverless motorcycle (with a sealed passenger compartment) is what you’ll call to take you to worka vehicle that could very easily get the equivalent of 150-200mpg. If you’re transporting your child to a doctor’s appointment, a 2-seat roadster sized car. If you’re taking the whole family into the mountains for the weekend, a full suburban could show up. But the point remains, most vehicles owned today carry 75-85% excess space and weightbut we need to lug it around cause we can’t effectively match our personal vehicle fleet of 1 or 2 to our changing trip requirements. While that’s true the USA has seen a ‘barbell’ employment market replace the post-war boom years of ‘well paid work for all who want it’.

Low-education medium-skill factory jobs in manufacturing sector and same for administrative jobs with software wiping out entire career paths. This has greatly diminished the middle wage earning demographic. So at one end a need for lots of low paid service industry jobs doing other peoples chores for them (low paid because USA has comparative pathetic minimum wage and job security compared to OECD) and at the other end very highly paid, tertiary educated, thought-intensive occupations in IT, medicine, law, finance and so on. I read about a place called the Jungle the other day in San Jose where there was a homeless shanty with a population of around 7,600 homeless people — much like you’d see in Bombay only a short walk from corporate HQs of Silicon Valley. One woman being evicted said she’d lived there “all her life”. Self-driving cars don’t rely solely on satellites. They also use internal maps, and a huge array of sensors.

The cars can detect hazards on the road and dynamically plan routes around them. Everything is super redundant – but in the event that the car’s sensors fail it can just park itself. The same cannot be said about drivers. Human error is the cause of most accidents. These cars will never get drunk, will never be tired or stressed out, and will never get distracted. I don’t know how they’re gonna solve the driving in snow problem though.

Interesting Article. A few odd points in it though.

I cannot fathom that the American fleet of cars will drop by 99%. It simply does not make sense, as such a small number of cars left could not handle peak needs – such as rush hour, or going to and from a big sports game. I think there is a way though, that owners could control things a little better – they buy an autonomous car, and use it for themselves. Then at times that they know they will not need it, they are able to put it into “rentable mode” with a management company – much like people that own condos but – for most of the weeks of the year, the condos are made available to rent to vacationers through a condo management company that handles rentals, payments, cleaning, etc. This is all very well, but how many times will the consumer of this autonomous service put up with the detritus left behind by those who have used the vehicle previously?

Litter and biological secretions of various types that will inevitably come to pass without the custodianship of a driver to keep all-too-common human behaviour in check. Also, it occurs to me that the only way this could really come to pass in the way you envisage will mean the institution of some kind of uber-dictatorship, enforcing the passage of the necessary rules and regulations required to ban traditional consumer vehicles from within city limits.

The democratic process is the enemy of this particular vision. There are many other holes I can see in this plan, to be honest.

It’s very idealised and the timeframe you foresee is extremely short. Maybe it will happen eventually, but not as quickly as you predict, I think, and it will meet with quite significant resistance from many politically influential demographic groups.

I think you make good points. I saw the one about being forced to (with no control once on the trip) put up with other passengers, right away. As a computer technician for my career who has studied and played with artificial intelligence, I also agree that there are quite a few barriers which will slow down widespread adoption. The good thing is, it will get a chance to prove itself. It won’t be widely adopted unless it DOES prove itself and is highly beneficial. That’s the great thing about a democratic capitalistic society — you get lots of choices.

For me, as long as I can own the car and control who rides in it and how it is maintained, then if the other factors are a net positive, I’m all for it. Especially if it solves my likely mobility issues as I age, within the next 20 years or so. (Again, choices — I see myself as better off if I have them). • Pingback: () •. Interesting article – thanks Zack. I noted only one other motorcycle comment and no response to it unless I missed it.

I will blame the font for that:-). Anyhow as an avid motorcyclist I wonder how we might be affected.

Would hate to be banned from the cities like my fellow car enthusiasts who like to drive their vehicle as much as I like riding mine. Reckon I will be joining Sons of Anarchy to get that sorted out. A self drive motorcycle won’t be the answer. On a more serious note though I sense some of my freedom of choice under threat if this comes about as a like it or lump it solution. The best technology solution is not necessarily the best from a sociological perspective. Enjoyed the article Zack, I follow a few blogs about autonomous vehicles so it was great to pick up a few new ideas.

A couple of thoughts, whenever traffic commute times get reduced people tend to apply for and get jobs further away from home. Before long the congestion returns. This is especially the case if car transport costs per mile are reduced. To lock in the benefits from this technology, some sort of long distance travel price penalty would be needed.

There may also be Uber/Taxi industry style demonstrations and strong corporate lobbying as the jobs and companies threatened start to fight back (and they will fight back!). Against this the environmental movements, which probably have not seen the environmental benefits yet, will join in the action. Perhaps some interesting times ahead. Because many people enjoy driving and electric vehicles will have limited range, many two car families will probably hang onto the last vehicle for a bit longer.

The remaining car would allow for long distance holidays and allow driving to become as more for recreation rather than a necessity. There will still be a need for concentrated parking/recharging areas as vehicles wait it out before the next peak hour. However these areas could have the cars packed in very tight as there would be no need for spaces for doors to open or cars to get out in any order. Cheap land near a substation could be a winner.

And car chases in movies are just not going to have the same excitement in the future. I really liked your analysis but think you might be a little off base on the environmental impact. You said ” Passenger cars, SUVs, pickup trucks, and minivans account for 17.6% of greenhouse gas emissions – a 90% reduction of vehicles in operation would reduce our overall emissions by 15.9%.” This 17.6% assumes the vehicles are being driven 4% of the time.

When we have a 90% reduction of vehicles then the about of drive time is going to increase significantly there by increasing greenhouse gas emissions. This will be mitigated by electric vehicles, but we should expect a spike in the short term when there are a large number of autonomous gas powered vehicles.

Trucking is going to continue to be diesel driven for a long period of time. Fantastic summary. One thing was missed.

My city used to span 30 blocks. The city stopped at the end of the street car tracks. When the automobile increased mobility, the city quickly sprawled out over 150 blocks. I think the same will happen here. Urban areas will be easier to live in without needing parking, but suburban sprawl will be the biggest change as anything within 100 miles of a city will be within the daily commute. This will increase the amount of available housing and lower our #1 cost, housing, •. “Tesla Motor’s CEO, says that their 2015 models will be able to self-drive 90 percent of the time.” 90% of the time a typical driver uses it or in 90% of the places cars go every day.

Driving on highways is easy, city streets not so much. “which is an astonishing waste considering that the average cost of car ownership is nearly $9,000 per year.” Which includes fuel and maintenance, which are dependent on miles driven/riden which will not go away with autonomous ride sharing. “90% reduction of vehicles in operation would reduce our overall emissions by 15.9%. As most autonomous cars are likely to be electric, we would virtually eliminate the 134 billion of gasoline used each year in the US alone.” Same number of miles driven/riden means the same greenhouse gas emissions. It would just be 1 car doing 100k miles a year rather than 10 cars each doing 10k miles a year. “PricewaterhouseCoopers predicts that the number of vehicles on the road will be reduced by 99%, estimating that the fleet will fall from 245 million to just 2.4 million vehicles.” If everyone is going to work at 9am, and there are 100 million workers, how do 2.4 million cars deliver them all at their work place at the same time? What about farm vehicles, autonomous or not, being 8 hours from a city would keep them out of the daily rush hour commute.

Great article. For a web simulation of how a fleet of autonomous cars operates in a medium-sized but low density city, Canberra Australia, see. The direct savings come from higher utilisation of expensive assets.

The congestion reduction comes from rising average vehicle occupancy in peak periods from 1.1 – 1.2 commuters to 1.8 – 2.2. Further congestion reduction comes from better traffic coordination and “platooning” (not modelled in the simulation).

Indirect savings (also not modelled) come from higher value land use, shorter commutes, more productive use of travelling time, fewer accidents and greater mobility for the entire population. While there will be less cars, and more efficient maintenance; things wear out. They wear out when they are used not as they age.

So if a new autonomous car is driving 100k a month it will need a lot more things replaced than a people driven car doing 1k a month. At some point, averaging around 200-400k current cars just stop, it is no longer worth replacing parts is no longer worth it and a new car needs to be purchased.

In a world with only 3 million self driving cars, the auto industry might be able to stay afloat a while if those cars need to be completely replaced 4 times a year. Thanks for writing this. I feel driverless cars will have a huge impact in making life easier in metropolitan cities.

These could also help make public transport viable in smaller cities with less population density. Cheap and efficient driverless cars can help solve the last mile problem and connect commuters to public transport. Wind Wifi Wpa Calculator Infostrada.

The real problems that will have to be tackled are safety standards – if driverless cars are required to implement the same safety standards as gasoline cars and electric cars then they will not look like the Google toy car and will cost not less than current cars. As long as driverless cars have to share highways with regular cars they will have to follow the same safety regulations.

Those of us who are in the technology industry and are early adopters of trends want to see transformation happening really fast – but in the real world things are ‘used’ as long as they can be ‘used’. The millions of automobiles that are being produced right now will start going go off the roads after 2030. This is also going to be an existential crisis for men – what are men going to do? What will they talk about? And please don’t say more workI hope space buggies come about before these driverless cars take over.

A company such as Uber (or a car manufacturer seeking vertical integration) running a transport utility, or a cooperative of individual owners putting their car into “hire” mode as you suggest, or a public utility are all possibilities. Maybe there are economies of scale for cleaning, handling complaints and lost property, booking, payment, predictive fleet deployment, break-downs, automated recharging/refuelling that will advantage a fleet coordinator, if not a fleet owner. Maybe there a “public service” obligations as well to provide transport to the disadvantaged which could require at least some government involvement or subsidy for some journeys. More here: •. Yep great I agree. I guess I just think of current electric cars like Tesla as a sustaining innovation for automotive industry.

BMW and others have electric cars already. These companies don’t sell drivers so for now at least the pressure of disruption will be elsewhere (on the industries that provide drivers). Those that sell cars, will need to change and I agree the numbers you suggest would make it hard to survive for sure. But I am still not convinced Telsa is any different from BMW, the form is the same. It will become disruptive tech for car manufacturers when a new form of car emerges.

Like a $400 car built and designed by IKEA or Herman Miller or Apple or something like that. And where the car looks nothing like a car, but instead like an office, a lounge room a gym etc, as well as smaller cars that do only specific things like deliver food etc probably bundled with software services. And in respect to number, these new cars (that don’t resemble cars at all) will proliferate massively. Numbers will increase, there will be more of them on the road and traditional car manufacturing will seem small in comparison. It’s just that they will come in a range of new forms (from small ‘cars’ that drive your bags to the airport for you, or you laundry home etc), to bigger ones that drive you to a destination while you run on a treadmill inside it like a mobile gym, or a mobile office etc.

And just like how the amount of computers has increased from a few mainframes to all these new forms like tablets and smartphone, and how these have become almost ubiquitous, where everyone has a smartphone, and millions of apps, and it has become a utility I think the same will happen with these new auto transfer vehicles. The form of the car will change. And that will be the challenge of disruption for all car manufacturers including tesla.

Dosi does a good job of this too Thanks for the thought provoking read. Just to add- What I am saying is that autonomous cars might be disruptive to Uber just as much or more than for automobile companies. Uber are a driver company. Changing themselves into a non driver company won’t be easy (even if they say it will be- sure they have consumer, but so did Sony with the Walkman), and there will be a rise of alternatives to offer the same thing, perhaps better bundled in other services or products (rather than just a peer to peer riding service). This makes it look like the future, but may seem totally pointless later (like mini disc players, or digital casette tapes which launched at the start of the digital revolution in music, and for a while seemed like the future but ultimately were a red herring). Not to say Uber can’t do it, but, I think there is still so much to play out, they got to get a lot right, and like you I think its all going to happen fast. There should not be a dramatic drop (one order of magnitude or more) in sales.

Vehicles that were being used 4% of the time (from article) would be used close to 100% of the time, meaning they will wear out about 25 times as fast. Consumers are already holding on to their vehicles for eight years on average, the point at which many major parts and systems need costly maintenance / replacement. While the large auto makers will need to switch to a model that is more oriented toward producing fleet cars and parts for maintenance, I don’t see a significant impact on overall manufacturing rates. “Morgan Stanley estimates that a 90% reduction in crashes” Yeah, right.

And flying pigs will deliver that. When automatic car has control of 90% of the time, humans lose control few times in an year, in any conditions.

That’s few seconds out of 8760 hours/31 536 000 seconds. And doesn’t even apply to _all drivers_. Or 99,99999%, a bit more. Let’s return to this when automatic cars have _even the same reliability than humans_. With 90% reliabliity you will get a crash in less than an year, for every car, every year.

How will this “reduce crashes” and not increase them? We dont’t, yup (3 oldies, >25y) But then the big money steps into picture and forbids private car owning and/or driving. Either you rent a bubble or don’t move: Using laws to ensure profitable business is a common business model now and every day. See what Hollywood is doing: They are very good at this. I see the reliability, cost and usage are basically forgotten and huge amount of money is wasted while a fraction of it could have made current car last 10 times longer than they do now.

It’s not even a technological problem but business problem: Manufacturing cheap crap creates more profits. And waste, but who cares about that? So the waste -argument is mostly moot: Any current car could last at least 30 years of daily usage, up to 400 000 miles, if manufacturer wanted that. Why would anyone believe they would want any better autonomous car?

They don’t: It’s all about money: A lot of cheap units bring in more money than few expensive units, unless your units cost billions per piece. Reliability, both for actual driving and lasting for a long time are far from what reality would need: At 90% you will have a crash within an year, for every car and every year. A deathtrap: These bubbles would have zero stars in a crash test if someone dared to put one in. Cost: Who will pay these obviously ultra-expensive cars? And who will maintain them? It’s obvious that claimed $0.50 per mile won’t be a reality in tens of years when a (current) car costs $1M and lasts 50 000 miles. If even that, Google isn’t saying.

Usage: 36 seconds in NY? You’d need to have a free car in every block to achieve that on average. Afaik NY has a bit more than 9000 blocks and it’s possible that some of these cars are actually in use. Also: Who cleans these 9000 cars? Every one of them will look like a pigsty after a day in NY: Dirt and puke everywhere.

How clean taxis are and those have a driver to watch it and clean it, all the time? A cost which doesn’t even exist in any calculations presented here or anywhere: None of them have any basis on reality whatsoever.

Then the legalese: Who do you sue when you have a crash in an automatic car? Class-action suits will be flying like flies on a ripe banana on summer. Summary: Over-optimistic forecasts up to hype level with both feet firmly in the air. None of that will happen until the legal problems are solved and that alone might take tens of years. I see that an autonomous car can replace taxis and possibly buses, but as a community based service they won’t replace private car ever. As a private car, possibly, but for that it needs cost to be comparable with current cars in every segment: Purchase, maintenance, insurance and fuel.

I can’t see that happening until the technology is very mature and that will take a long time. Unless big money buys legislation to forbid private driving and no no, profits won’t be the explanation, but of course old lie: Safety. Even with 90% reliability as “90%” looks like a high number. Then someone like me does the math: “This is safe 330 days in an year” then it doesn’t look safe at all.

But once the law is there, it’s irrelevant if the safety is ever reached, no-one ever checks. But you can bet the profits will be there. Number of vehicles on the road will be reduced by 99%, estimating that the fleet will fall from 245 million to just 2.4 million vehicles” A number shaked out of the sleeve based on absurd assumptions. We have more than 20million people living in rural areas and calling a taxi (=autonomous community car) to somewhere in the middle of nowhere would take hours, so no, 2.4 million vehicles is an totally absurd number. 24 millions might make sense if you forbid private car ownership by law but somehow I won’t see that happening either. Summarum: A consulting house is paid to provide credibility to made-up numbers. Seems to be the standard procedure nowadays, everywhere.

I’ll bet a good bottle of whiskey that it won’t be even legal to have an autonomous car driven for you on public roads without you being the driver in legal sense and thus responsible for _anything_ by 2025. Especially speeding. I call a car “autonomous” when I go and sit on the back seat and if something happens, it’s the manufacturer or, in any case, _somebody else than me_ who is the guilty part. For anything, including speeding.

And I won’t believe that will happen in 10 years, just 2 elections away. For technological, economical and legal reasons.

Finally, thank you Thomas, for injecting some sanity into this mental masturbation of zero car ownership and a world without the automobile as we know it today. Thomas brings up a very good and “sticky” point, “We have more than 20million people living in rural areas”, it begs the question, how are “they” going to deal with that? Are “they” going to force them to move into the big cities?! Is technology going to override the laws of physics?!

Those 20 million people, that’s 20 million votes (more if you include relatives) in an election year, chew on that. “it is unlikely that major automakers like General Motors, Ford, and Toyota will survive the leap. I think that most will be bankrupt by 2030” Assuming that quote really predicted the future, what can fill the void in the global economy IF just those 3 auto manufacturers go under? Is some sort of silicon valley wet dream going to even make a dent in that economic crater?!

“The Bureau of Labor Statistics lists that 884,000 people are employed in motor vehicles and parts manufacturing, and an additional 3.02 million in the dealer and maintenance network.22 Truck, bus, delivery, and taxi drivers account for nearly 6 million professional driving jobs. Virtually all of these 10 million jobs will be eliminated within 10-15 years, and this list is by no means exhaustive.” There some people out there who are rubbing their hands in glee over the above quote, weather out of malevolence or just plain lazy ignorance. IMHO, I don’t think the US economy can withstand a loss of 10 million jobs within a 10-15 year span, hmm unless the REAL goal IS the destruction of the USA as we know it now it makes more sense. The whole article has the “smell” of a globalists’ pipe-dream. “Assuming that quote really predicted the future, what can fill the void in the global economy IF just those 3 auto manufacturers go under? Is some sort of silicon valley wet dream going to even make a dent in that economic crater?!” Tesla.

“There some people out there who are rubbing their hands in glee over the above quote, weather out of malevolence or just plain lazy ignorance.” For the record, I own a brand of performance auto parts. If what I write turns out to be true, it would decimate my industry. I’d say that makes me fairly objective. 1) 90% Autonomous Cars: The existing technology of self-driving cars *currently* has control of the car 90% of the time. As you mentioned, this is clearly not ready for consumer use on a mass scale and still requires drivers who know how to drive. They’re predicting that it’ll be “fully-autonomous” or self-driving 100% of the time in the near future, which makes it a different story. 2) Crashes: We don’t have any data on how often crashes happen so its tough to speak to this.

I did read an article a while back about Google, however, about testing “aggression level” inputs for cars, which makes them either super aggressive and prone to crashes, to very cautious cars that prioritize the safety of both cars and pedestrians (e.g. If it looks like a pedestrian might make a cross, it’ll stall for a few seconds while the pedestrian makes up his mind to cross. I imagine there’ll be some sort of signalling added to cars to simulate what pedestrians currently use to figure out whether or not to go, which is eye-contact). 3) Cheap Manufacturing: This one is a bit specific to currently existing companies, but if Tesla Motors is the predominant supplier of autonomous cars, I imagine they’ll be more interested in providing the highest engineering standards, just because its Elon Musk–a man who is arguably an engineer before he is a businessman (case in point, he lets anyone have access to his technology in the belief that it’ll accelerate positive technological change). Either way, it’ll still be profitable with a dominance in complementary industries such as electric stations (unless it becomes a grid-based car system), maintenance stations/services to industrial buyers of the cars, software support, etc.

Even if the cars last a long time. 4) Public Transit: I think autonomous taxis wont replace public transit like buses, trains, etc. As long as they’re cheaper or faster.

5) Legal issues: I think this is the most important point brought-up by Thomas as an impediment to the implementation of the technology. I think San Francisco perhaps will push for municipal legislation to speed things up as a pilot city after some smaller scale trial outs of the technology. But even then, it’ll probably take many years, so I’m also not optimistic about the ten-year timeline. It’d be interesting if Zack could maybe get in touch with some tech lawyers from the Valley and interview them on their thoughts about it? Seems like they have no more access to this technology than anyone else (outside of investment from Google).

Seems like the barrier to starting a really great/cheap taxi service is going to be greatly lowered. And all the durable advantages Uber has built so far will be mostly useless. From what I see, it doesn’t look great to be Uber. They may have the vision, but not the power to execute it alone. The real winners will likely be whoever is licensing out the self driving technology. Maybe that will be everyone.

Right now it certainly seems like Google has a leg up. Zack, why say things like “Hertz pays a lot less for a car” with NO citations?

With respect, it appears you are making an unwarranted assumption. In the sedan market, profit margins are razor thin – roughly 1%. The car companies feel compelled to compete there due to the sheer number sold, but they don’t make much per car. The overall profit margin of car companies tends to be roughly 5%. So, if you think Hertz is buying cars appropriate for taxis for 50% or even 20% discounts en masse, you’re delusional. You can Google search terms like ” sedan price margins” and get lots of figures — don’t take my word for it.

Here’s an article on tight sedan margins I just read on Bloomberg the other day: •. It’s definitely an interesting future to ponder. But what about Uber jacking up prices once it has considerable market power, thereby making the economics less attractive to some consumers? Also, I’d be interested to see analysis on the average commute (non-public transport) in the US. Outside of urban centers (where most of Americans live), people drive to work, drive to the grocery store, drive to the mall, drive their kids to school, drive to restaurants, and each of those could be 10+ miles. Does a taxi service really make sense then for all your needs?

“The total design of cars will change.” I doubt that. Current ones are based on safety, comfort and luxury. And profits for manufacurer, of course. Leave any of them out and you won’t have much paying users, even with cheap price as price isn’t only selling point. Plastic seat with puke on it? No thanks, even for free.

Also: I can see that current autonomous cars would get zero stars in a crash test so they have to change as there will be crashes. Software can’t predict the future or beat the laws of the physics. Even less than a driver can. “The environmental impact of autonomous cars has the potential to reverse the trend of global warming and drastically reduce our dependence on fossil fuels. Passenger cars, SUVs, pickup trucks, and minivans account for 17.6% of greenhouse gas emissions26 – a 90% reduction of vehicles in operation would reduce our overall emissions by 15.9%” This doesn’t make any sense. It’s number the number of vehicles in operation which drives consumption, but the amount of operations.

If people are traveling 20,000 miles in 10 cars or 20,000 in 1 cars, the output emissions are the same. Now there of course be other reductions, but what was stated is reduction ” vehicles in operation”, unless the 17.6% is from vehicle PRODUCTION (it’s not). Furthermore, there’s the distinct possiblity that making drivng easier/less taxing could INCREASE the amount people are willing to travel. Thus increasing daily emissions output.

“electric cars could actually increase green house gas emissions” Electric cars, like the Tesla, are way more efficient in terms of MPG than ICE cars. So even running on fossil fuel sources they still pollute the environment less than ICE cars.

Not only that, there is nothing that you can do to increase the efficiency of an ICE car; but with an electric car you can install solar panels (or any other renewable source of energy) and the pollution that your car contributes to the environment rapidly drops to zero. You are never going to get zero pollution with ICE cars; electric cars win here without a doubt. In short, that “could” is just not true; electric cars are already a net win for the environment. After all, it’s not like we are comparing an imaginary electric car to a real ICE car. We have real electric cars today and they are already much better than ICE cars and the gap will only widen in the electric cars favour.

Once all self drive cars are replaced cars could get a little lighter, but probably not much. Driving bumper to bumper doesn’t make them more efficient, not having to stop and slow down does. You would certainly gain some efficiency by having cars not stop at intersections or slow down when someone ahead decided to slow down for a turn without signalling. City fuel efficiency would probably become higher than highway, as the speeds are lower.

Being electric will not necessarily increase efficiency. Electricity needs to be produced in some way somewhere.

Charging a battery and then discharging it also lowers the efficiency further. From a green house gas perspective, burning gasoline or natural gas on board a vehicle is far more efficient than burning coal or natural gas at a power plan, transporting it through high voltage lines to a charging station, charging a battery, and then discharging a battery.

I live in Austin, and have been thinking a lot about how we can improve transportation in this city. It is refreshing to see others searching for innovative approaches to avoid the massive, costly infrastructure investments in public transportation—that will barely put a dent in our traffic woes. How do you see larger vehicles (intra-city buses and mini-buses shuttling groups on high demand routes) fitting into this picture. Even right now, especially in Austin where public transport is minimal to non-existent, I feel there is space for a hybrid uber/minibus network, that brings an innovative approach to moving people across the city. Thanks for your thoughts, really appreciated it and am sharing eagerly. Good article. I especially appreciate you points about reducing death, injuries, and economic damage associated with human-operated vehicles.

I’m not sure I agree that this technology is the biggest innovation we will experience. Societies still learning how to deal with the telegraph (mass communications), computers, and other mobility technologies, like flight. Cars have not changed much in 100 years. Autonomous cars are more the convergence of other technologies (electric cars, computation, AI), then something new. It’s also important to remember that we have a say in the future, which it’s why it’s important to consider your predictions. Great Post Zack, What I would like to see is your opinion from an investors point of view.

If you would manage your own portfolio. What stocks, would you stay away from and what stocks would benefit form the SDC transition. Like you mentioned there are a lot of industries that will be affected by t and probably much faster then anyone can expect it. What I also find interesting is the emerging of two groundbreaking technologies in the automotive sector in the same time. SDC technologies 2. Affordable EVs. With the SDC fleet theory, I see a massive acceleration in adaption of EVs then previously predicted by the industry.

Which in turn offers other opportunities to invest. Other areas I see affected by SDC is long distance public transport. Taking a SDC for A to B on a 200 miles trip would be the same price but more convenient than taking a train.

Im looking into MobileEye right now, but its hard to determine how dominant the will be and how Google holds up against them. Their CTO stated that he think cars have to be connected to be fully autonomous, I don’t see that as a viable solution. We will wait decades till every car is connected, a solution without the necessity of connected cars will win. One thing that doesn’t seem to be taken into account is the human need for individuality and a feeling of freedom, and the desire for banks to lend money for people to realise that dream. I guess that as long as you are only talking about utilitarian vehicles in cities and out on arterial routes then a majority of these will eventually be changed to autonomous vehicles, but don’t underestimate mankind’s desire to have their fate in their own hands. As soon as autonomous becomes cheap enough, people will want to own their own, regardless of environmental impact.

My view is that there will be a cut in vehicles, but that it will be a slow decline followed by a quick recovery as more people can afford their own cars (either purchased on finance or on a long lease). I don’t think this will happen nearly as quickly or act as efficiently as you write. While everything you say leads to logical conclusions, things like the decommissioning of curb side parking will be a municipal decision, and anything involved with the democratic process will take a very long time to be pushed through. Another point of dispute is the availability of auto-cars. I believe there will be levels of auto-cars. There will be basic models which have nothing more than a bench and move from point A to B, but then there will be cars with extra leg space, cars with desks, cars with wifi, cars that can remotely and securely access an individual’s work data.

All of these things will cost a premium. Some people won’t care and will go for the cheapest option, and some will and they will likely have to wait for one of the fancier cars to swing around. Eventually, all of the auto-cars will come with these options, but the cheaper ones will have to go through their life cycle int he fleet. My point here is largely, the level of efficiency you claim won’t exist as quickly as you say. Will everything you write about happen in your and my lifetime: absolutely. Will it happen in the next 10 years?

I’m doubtful. The next five? Cars are parked most of the time because people work around the same time and people sleep around the same time (also shop, and do other stuff around the same time). Unless we stopped commuting altogether and just stay home, this is just crazy talk.

You need MASS transportation in the morning and MASS trasportation in the afternoon, so unless 99% of people stop commuting to work this is just silly to even discuss. Most people in urban areas already have rideshare options, the fact that now they’ll be driven by “no one” doesn’t change that fact. This would only be true if everybody who wanted a ride was willing to be queued up in the world’s longest line.

The reality is that most people, especially in urban areas (where this would be implemented), commute at the same exact times. Your fleet would need to handle the sum of all commuters in parallel, otherwise your new transportation utopia would look more like a regression of service, and these cars wouldn’t catch on. There is some batching that auto-cars could do, but no more than a regular taxi or ridesharing service. Basically, you need to take a picture of rush hour traffic and have that many cars available for that area, if not more. No, with respect, you’re missing the point that the disagreement a LOT of commenters have is not that the car fleet will be reduced in size, but that it will be reduced anything REMOTELY like 99%. If you actually read comments like Paul’s and think about them, lots of good points get made. Empty cars driving themselves to their next destination, for example, will be miles driven that aren’t even driven today — and that will actually add to congestion.

A key point I haven’t seen made much is one reason people don’t like riding the bus is the random other passengers, which may be unpleasant or even dangerous. I think the idea that everyone will be happy to share a ride with several other RANDOM passengers — and no control over that situation — are completely unrealistic. Thanks Zack for a well-researched article. I hope things play out exactly as you predict.

I think tsc444 gets at the key concern, though. The reduction in the fleet comes from removing cars that can be replaced by others that aren’t currently being driven—basically minimizing parked “downtime.” People still need to go wherever they go today, so the amount of traffic doesn’t decrease. The reduction in the fleet may be overstated, too. The question is how many cars you need at peak demand, during rush hour. Is that really only 1% of the fleet, that is, are 99% of cars really parked during rush hour? Another thing to consider: With autonomous cars, the number of vehicles on the road can be more than the number of drivers.

Other things equal, empty cars driving themselves to their next fare increase, not decrease, congestion. Broadly I agree—and hope—that it’ll be economically nuts for the average person to own a car in another 10–20 years. Just not sure if the traffic prediction holds water.

Nice foresight, but you are missing three factors that will no doubt complicate this future. First, rush hour. When two million New Yorkers all leave work at 5pm, that 38 second prediction will be off by two hours. Second, to deal with rush hour, we’ll likely see autonomous vans and buses, sharing ride sharing vehicles. Third, a large number of Americans can’t afford $0.50/mi.

The cities will no doubt tax ride sharing to keep public transit running, initially for the 99%, then 90% and onward down to the 50%. Great study, but flip to pages 41-43 and you find a major flaw, one which will lead to far higher wait times and increased costs.

The model they use assumes an even distribution of riders across a city. Here in Seattle, we can see how that mismatches reality when looking at the distribution of car2go cars across a typical day. The cars sleep at night out in the exurbs, and flood downtown all day. Thus not only does this model not take into account the massive flows of riders in the morning and evening rush hours, it assumes almost twice the capacity of the network as the cars do not dead head back out to the Ann Arbor burbs, or to Brooklyn, Queens, and NJ to pick up the next person in the morning. At $1/mi in Ann Arbor, the conclusion is the same, but the arguments about busy mass transit in Manhattan etc. Are far less convincing when I have to reserve a car 5-10 minutes from now vs.

Walking to the curb and waiting 26 seconds. Luni: my initial reaction to the AnnArbor study was the same as yours when thinking about my home town, Canberra Australia, which has highly “tidal” flows to/from far-flung suburbs. So I wrote a web based simulation of Canberra’s traffic and how a fleet of autonomous cars would handle it. You can vary the fleet size to get different waiting times: In summary, a fleet of 23,000 4-seater vehicles can service 750,000 journeys a day (from a population of just under 400,000) with 98% of cars arriving within 1 minute of being requested, as long as car sharing is used in peak periods for commuters travelling from nearby sources to nearby destinations. Total (capital plus operating) costs per mile are less than half those of private car travel. I’m pretty exited about not owning a car in the future.

One thing I wonder about the statistic about cars only being used 4% of the time, though, is that it seems like that’s not evenly distributed throughout the day. In other words, for a given car the specific hours that it spends on the road are likely to coincide with the specific hours that most other cats are on the road (I.e. Commuting hours). That doesn’t seem to square with the prediction that 2.4 million autonomous vehicles will replace both the current fleet of 245 million vehicles and all public transport. Unless commuters display a sudden change of behavior and start carpooling or move closer to work it seems likely that we’d still see large numbers of vehicles on the the road. Holy moly Zack!

Amazing stuff. Thanks for putting this all together: made for an exciting read. I missed the self driving car demos at CES this year, but I heard people raving about them. It seems like a part of this system that needs to exist is a standard communication protocol for self driving cars to communicate with one another. Another is a technician network (nerds in cars) to show up when the transmitter in my self driving car stops working. I can imagine chartering a cross country limo with an office on board instead of flying for trips under 500 miles.

Amazing stuff. Thanks, Will. I am not sure that a car-to-car communication system is necessary, or even ideal. This would be massively complex and would introduce many points of failure. I think that a distributed system where each car is making its own, independent decisions is probably a better design. I could see data from car-to-car communication being an additional and valuable data point, though. Regarding technicians – I think that autonomous cars will be serviced in large, centralized service centers, rather than having a technician come to the car.

They will automatically schedule themselves for service. If a car breaks down unexpectedly, an autonomous tow truck will be sent to fetch it. I think Hertz’s rental business is going to die, but it’s possible that they have a future in car maintenance – they have a lot of expertise in maintaining vehicle fleets. I think you bring up a great point about ‘local’ trips (.