Ask HN: What will self driving cars do to real estate? | Hacker News Reader

Ask HN: What will self driving cars do to real estate?

66 pointsoctaveguin79 comments

It seems inevitable that self driving cars are coming. When they do, they will change housing dynamics in a fundamental way.

Will this change make city living more or less desirable?

A few forces could be at work here:

* Parking lots in major cities are free'd up because they no longer make sense creating more housing real estate. City living is more affordable.

* Traffic Congestion is reduced so commutes are shorter creating more demand for suburb living.

* Distances feel shorter because they are no longer inconvenient so living outside the city makes more sense - more suburb demand.

* City living gets even more desirable because of the abundance of cheap self-driving cabs not available in more rural areas - city prices go up.

* More?

I'm sure some of these things will have a very little impact while others will have huge impact. I just wonder about the net result. Any speculation?

View on HN

I see two opposing forces:

Self driving cars could encourage sprawl by making commuting long distances more practical. If you could be working during your commute instead of driving, and didn't need to find parking once you arrive, a 2-hour commute could become practical. That would intensify the "car culture" that currently drives municipal planning, but would reduce the harmful effects. As you mention, less space wasted on parking, less congestion etc.

On the other hand, driverless cars will be a big boost for Uber, Lyft, Zipcar, GetAround and the like. That could make city living more desirable, because car-on-demand only works well when there are lots of available cars nearby. Folks that live out in the boonies would still have to own their own cars, and so wouldn't benefit as much from the technology as city-dwellers. Also, spending less space on parking has a bigger effect in the city where space is already tight, so we might see even higher density in the inner city.

I think that's about right. To your "on the other hand" though, I'm inclined to think that where density is already high, robo-driving doesn't change the equation of car + $10-20/hr driver all that much. (Especially given that some of those costs--cleaning, etc.--have to be shifted to a non-driver.) Therefore, I expect easier driving = more driving (which is more or less basic economics.)

Some no frills car sharing companies make their users clean the car. And if you get a dirty car you can order a new one and the previous driver gets billed. And small one person Robo cars could also be build much simpler and cheaper.

Correct, you'll be rating the previous passenger in your Uber's once we have self driving vehicles. Same mechanics will apply.

There could well be discriminatory billing - clean conscientious passengers pay less than drunk party goers.

Sure. My basic point though is that, given high utilization, $15/hour doesn't really change the economics of ordering a car on demand though. It still makes sense in the cities where it already makes sense unless services like Uber are already being blocked for regulatory reasons.

Cities where Uber and the like are being blocked with oppose robo-cars even more vigorously. Like the revolts against steam engines and factories we read about. Uber-opposition will seem like a practice exercise.

While lot of people are thinking about how the self driving cars will effect commute specially to work. We really don't know how the workplace is going to change. I think the conventional workplace is ripe for change. Will there will be more need to commute to work as opposed to more remote working with say virtual workplace with better virtual reality. Obviously some of the jobs can't be done unless you are physically present, but I am would like to think these jobs will get lesser and lesser. If I have to go to workplace lesser, I as well might leave far off from workplace. Also with self-driving cars will be lot as stressful as driving. I would assume self-driving cars will give us the chance to work farther from workplace. And less dependent on near to prime location.

Programmers are the prime example for a job that can be done remotely, and yet there are tons of articles about the problems it causes. The problem of staying in contact with your workplace hardly can be solved with virtual reality.

Remote work can only function if the teams have both the communications tooling to support it and the team and organisational culture aspects to keep it up. I find most fail at the latter, because they don't know how to be sufficient enough in communication with the tools they have. Talking to face the face has much more communication bandwidth than anything else out there.

Personally I would have expected to see a 4 or 5 times imporvement n productivity from not being in an office where I was interupted maybe once an hour on average. Saying that a more competent manager could have probably elimiated a lot of that (instead he caused way more interuptions).

Study after study says otherwise. For various reasons, small teams working in a huddle together for short sprints is the most effective way to organize a development team.

I was a team of one constantly getting pestered by end users who were to stupid to fill out an excel sheet properly. Or incompetent managers who were completely unorganized.

At startup size its considered unsolved. At multinational megacorp level its considered a solved problem, often enough your boss and some coworkers aren't even on the same continent.

One interesting social effect of self driving cars is I only see my bosses boss about every three months because he lives and works several hundred miles away, and I have coworkers I've never met in person. (note, not a remote job, just a typical "dozens of billions of dollars" non-tech mega corporation). Self driving cars mean I'd probably be stuck sitting in a car while working for an entire day to meet people face to face. For people who still believe in mystical laying on of hands, self driving cars would likely explode the amount of it that goes on, as I've found it weirdly true that famous projects are international, but its fractally similar in that you're also more likely to have a boss thats 30 miles away, or a bosses boss 800 miles away, not the proverbial instantaneous jump from the entire company in one skyscraper to fully distributed. Yes, a self driving car would be useless if my boss was in China, but things haven't spread out geographically that fast.

Life like virtual office with occulus rift

Why would self-driving cars reduce traffic congestion? If my car drove itself, I'd be less incentivized to use public transportation. I'd probably be willing to put up with a slightly longer door-to-door time than if I used public transportation because it would be more convenient: I'd rather surf the web in my own car than have to pay attention to stops, schedules, transfers, etc. Congestion would probably be different: it would be less stop-and-go and more uniformly slow speeds.

Because they could be better drivers. For example a "perfect" driver communicating with one in front of him could, eventually, tailgate with an inch at much higher speeds. If we approximate (for arguments sake) traffic as a fluid, by bernoulli's principal there will be more capacity for a given lane.

Said in another way, if the car can get there quickly it'll be off the streets sooner and another one will take its spot.

Also a "perfect" driver wouldn't make the stupid mistakes ppl do all the time, the most annoying one staring at gory accidents. And even accidents wouldn't occur as often.

But this all assumes that the AI gets even more sophisticated.

I guess it all depends on how much more capacity self-driving cars produce AND the increase in demand for self-driving cars.

I'm more in the camp of the GP. I think a big reason why people live close to work, take public transit and go without a car are all the things the GP described. Take those away and suddenly using your self-driving car to commute doesn't sound so bad. That equals a HUGE increase in car demand.

You're covering the available efficiencies in the use of existing road network capacity, and assuming that demand is static.

In fact, I think that if car ownership is cheaper, journeys can be made by anybody, and the experience is less stressful, many more people will travel by car.

I wouldn't be at all surprised if autonomous cars produce even worse congestion than today.

They will be networked, making demand pricing easy. So congestion becomes a political choice (if people refuse demand tolls in the name of fairness...).

Well that's the theory. What ends up happening is always up in the air.

I did read somewhere that the average "commute" has stayed constant throughout history - about 45 min - be it walking, omnibus, trolley or car. So you're probably right, things will shift to fill up the free time

People that are too defensive on transit tend to cause congestion, not avoid it. Of course, that's dependent of the streets being full of people a bit more violent, but until we have all cars driving themselves, that condition will be true, what may take decades.

Also, if the car is self-driving, people will be want to make more trips. For example, if we send their cars back home once they got to work, we'll save on parking, but get more congestion. I'd love to automatically send my car to a mechanic when needed, what's also more congestion... I'm sure people will find lots of good reasons for sending their cars around.

Anyway, what will probably reduce congestion is a smart bus system. Imagine a small bus that takes you door to door, and takes only slightly more time to get there than your own car, but is way cheaper. I've seen that on the news just recently, I think it's Japan trying it.

>People that are too defensive on transit tend to cause congestion, not avoid it.

Have you any proof of that? I would guess that both "agressive" and "defensive" drivers cause problems in different ways.

For example when I am waiting at a pedestrian crossing to cross, if a driver comes up fast and brakes hard close to the stop I won't start walking until I am sure he is stopping. If another driver slows down well in advance it indicates to me that he is inteding to stop, I'll start crossing earlier, and consequently he will start again earlier. I have also read articels about traffic waves being caused by people basically driving faster than the car in front then needing to slow/stop when they catch up with the one in front.

Idiots not moving from the middle lane must cause a lot of congestion in busy motorways, as they are effectively reducing 3 lanes to 2 for others around them who want to use the road properly (though I wouldn't class them as either defensive or agressive).

I like to characterize is as unpredictability being the real harm. If you don't behave as expected, everyone around you will have to react at short notice, and that disturbs the flow. And those disturbances cascade, because laminar traffic flow is highly unstable.

You missed my point: of course they will be more sophisticated, but if they increase the demand for road resources then there will be more cars out there. They will move efficiently but as slowly as people are able to tolerate before switching to alternate means (e.g. trains, biking, etc).

I agree it's not a given. It's not unlike building new roads which sometimes can increase traffic. You can get people who haven't been able or willing to commute so far join the commuters. If self-driving cars are shared they might be driving more miles and be more likely to break down and block traffic.

There is a possibility to optimize transport better using self driving cars but we may not actually take advantage of it. Even things like sharing self driving cars may not work in areas with larger sprawl because the probability of two people looking for the "same" ride within a certain time window and geographical area is small. In the city it's more likely but there are already solutions such as bicycles and public transport that work fairly well in dense areas.

It is very difficult to predict the impact. Perhaps impossible.

Well, here is something that recently occurred to me:

I'm a member of a typical 2 car suburban family. I work and my wife is a stay-at-home mom. Truthfully, I only need a car to get myself to and from work, while my wife needs a car for ferrying kids to and from school (and activities), shopping, errands, etc. Thus, we need two cars.

My car spends most of its day sitting in the parking lot at my office building, unneeded. However, if I can drive myself to work and then order my car to drive itself home for my wife to use (who would hopefully send the car to come and fetch me at the end of the workday), we could easily become a one car family. Which would be nice. Of course, I live about a half an hour from my workplace -- this might not work as well with a 2 hour commute.

If you own or hire SDCs, does your wife still ferry the kids around? Or do they ride alone to the destinations you've approved in the car's parental controls?

I expect the "Soccer Mom" occupation to disappear.

Well, with older children I might allow this. Young kids would need to be accompanied by a responsible human. Robots haven't earned that level of trust yet.

In any case, these technologies won't come along in time to help my current situation. I do hope SDCs become generally available in time to prevent an elderly jcadam from plowing into a farmer's market sometime in the 2050s/2060s. Hey, I guess 'senior mobility' would be another good application of SDCs.

From my perspective, the annoying thing about either driving or using public transport is it feels like wasted time. Sure, you can listen to some Great Courses while you're on a long drive, but apart from that you're trying carefully not to get in a crash. If you're on a subway or bus there's barely any space, so most kinds of work are out of the question. So the self-driving car will be a massive improvement because it frees up your time.

Net, everyone gains an hour or two each day.

Assuming the SDC is reasonably cheap to run:

- The school run changes significantly. You don't need to live quite so near school any more. Yes, there is such a thing as the school bus, but that ties everyone to the same schedule and some parents are not comfortable with sending smaller kids on it. Depends a lot on the culture where this SDC lands. (Where I live, kids walk themselves to school. I lived in places where you sent kids on public transport, too. Or a school bus.) In any case, it means you don't have to be with the kid as they go to school. And they can do whatever after school activity they like, it doesn't need coordination with public transport.

- You can get drunk anywhere, not just where there are cabs/trains/buses nearby. You can keep partying until you literally faint in your car.

- You can live quite far away from work. An SDC with a bed could be a comfy commute. You can get things done while commuting. Heck, someone will build a house where the bedroom moves (or just park in your garage, but someone will build it in). Brush your teeth, get into your car, wake up where you need to be.

- You don't even need a fixed address if your SDC is a camper van. Combine it with a remote job and you are on a constant tour. I reckon someone will do this reasonably soon and show us.

- Think of all the things that are limited by having to go home afterwards (clubbing), or are normally quite remote (eg race tracks). They can all be made more attractive if nobody feels they're wasting time getting to them. It's possible there will be extra rush hours during each day.

* For some families, it's no longer necessary to have two cars -- the car can drive itself back home for the other person(s) to use during the day. Or neighbors could even share cars.

* Centralized parking at the neighborhood level for cars to be out of sight and charging while not needed. Services could be built off of this concept to handle cleaning, maintenance, etc.

* Reduction in need for street parking in busy areas. Cars can drive themselves away to "orbit" or be parked while the owner is shopping/eating/etc. This allows for alternate uses for the space ... bike lanes, or just better pedestrian zones.

* Smaller lot sizes are possible for homes if driveways and large garages are not needed/downsized. Smaller parking lots for apartment/condos are also possible. This makes denser real estate developments possible, especially in suburban areas.

I would think real estate will become more spread out.

Personally, I wouldn't mind a 1.5 hour commute each way if I could code the whole time rather than drive, and if my car was electric.

If you ever choose to start a family, this number might not be as appealing. Just speaking from experience.

Paying a million dollars for a house in SF isn't an appealing number either, compared to $350k in Santa Rosa.

Less Space taken up by parked cars may mean more space for people to live. Though SF real estate will likely be insane for the foreseeable future any way you look at it.

That 350k might go up once we have self driving cars.

That was my original point.

the 3h can be included in the job time, so you only stay 5h at the office

When your kid's school calls and says he's vomiting and needs to go home, it will make more sense. You're still 2 hours away from home no matter how much work you slam out in the JohnnyCab.

This may derail the discussion but it's hard to discuss this question without first wondering the impact of self-driving cars to mass transit. At least in New York, being near a subway stop is a very nice amenity and commercial development typically flourishes around each stop regardless of neighborhood. If stepping into a self-driving car is as easy as walking 50 feet to the nearest subway stop, will real estate around subways command less of a premium?

I guess the bigger question is, will subways have as much demand once self-driving cars become big? Obviously, no, if the cost for hailing a self-driving car becomes almost as inexpensive as a subway. But there are things the city of New York could do to impact that, such as reduce street capacity (i.e. increasing pedestrian and bike capacity).

But that's just New York. Everywhere else in the U.S. where mass transit is not a daily option, automated cars will be a daily habit. Inevitably, as more and more people migrate to New York and expect automated cars to be the default option, it's hard to imagine subways (and buses) winning out, absent some major limitation in self-driving car production or capability.

New York will always be an anomaly...but yes, I'd have to agree with your speculation that suburban living would become much more attractive without having to worry about car ownership or driving. And this would most definitely impact commuters to New York. But if self-driving cars become a ubiquitous option, will there be the political appetite to continue subsidizing mass transit rail, such as the kind that is currently used to bring in upstate workers to the city today?

Lots of factors to look at in how this plays out in very dense and transit oriented cities like NY and large European cities. There is the potential for major problems with traffic as people move from taking transit to riding in a self-driving car. The subway infrastructure moves a lot of people, and if you move a lot of them up to the street level, the streets may not be able to cope. A lot of plusses and minuses come to mind.

1. Until all cars are self-driving, we will still need traffic lights, and so intersections will still be a big bottleneck.

2. Self driving cars may be able however to maintain closer following distances since they have quicker reaction times.

3. Jaywalking has the potential to cause huge traffic problems because pedestrians will know that self driving cars won't hit them.

4. Parking can be made more efficient by sharing cars. This is already happening in Berlin and other cities with floating carsharing systems.

5. Parking can also be made a lot more efficient if there are self-driving parking lots or areas marked off on the street, and the self-driving cars have a protocol to cooperate with parking. Parallel parked cars can be parked bumper-to-bumper, and they all just scoot out of the way to let someone out. In a lot, cars can park much more closely packed left-right since doors don't need to open, and the driving lanes can be reduced or eliminated if cars cooperate to get out of the way when one has to leave. This may free up street area to add extra driving lanes.

6. More people will probably share rides with a self driving Uber Pool/Lyft line type model.

7. If cars are shared, a lot of the fleet can be very small two or even one seat cars that take up less space in the road and parking.

8. Car interiors will resemble an office workstation. With mobile internet, very long commutes will be acceptable if the car is a fairly comfortable mobile office.

A royal commission did a study on "Traffic in Towns" in 1963 and concluded it would be impossible to serve a city with London-like size and density with cars even if you put the buildings up on stilts and built multiple levels of roads an highways -- the problem is that you can't support a high enough level of traffic on highway offramps.

New York City is pretty close to the maximum limit of car capacity it can support. For one thing there are only so many cars that can travel on so many lanes through bridges and tunnels. If you tried to add more lanes, you'd have the same kind of problems connecting those to surface streets that you get in the "Jetsons in London" scenario.

Robert Moses did have some plans for urban removal on the southern side of midtown to accomodate more traffic (and destroy traffic demand) but it didn't happen.

(Note the response to the car in most American cities was urban removal -- you can often find a big empty space in a second or third tier city that has a 'civic center' plopped down on it.)

Manhattan looks pretty much the same today with all the cabs that it would look like with self-driving cars.

The cars don't necessarily have to come back to surface streets just like the subways don't.

What if they redesigned the subway to be a bunch of self driving cars? basically have the stop be a lineup of cars that you get into on demand like a taxi and have each stop have a bypass so if your small car didn't need to stop there you could move on unimpeded.

I'm going to play the spoiler and say that I think everyone here is on the wrong track. In the short term, yes, I think it will business-as-usual, with some of the ideas being discussed here happening.

But think of the transition from horse to automobile. If you didn't predict the invention of the airplane, your guesses about transportation in the future would have been pretty far off.

I think the questions to ask are: -why will anyone commute to work in the future, when telecommuting becomes more and more lifelike? -what will "work" even look like? We can't be that far away from most service jobs being replaced by computers/robots. -Why would anyone go shopping, when anything they want can be delivered to them instantly or even printed on-site?

Yes but in the same vein we can't really predict why people will want to go from point A to point B in the future. Today the main reason is to go to work or go shopping but we may have much more free time and simple want to go to different places/activities.

By making automotive transport easier and more convenient, the obvious consequence is a general increase in automotive traffic.

In city centers the worst externality of cars is the space they hog. Self-driving electric cars won't fix that; instead they are likely to increase the utilization of that space: unlike with humans where congested traffic suggests the drivers to look for alternative options such as mass transit, an empty car doesn't mind waiting in traffic by itself.

Suburbia is likely to expand further because if you can work a few hours in your car that means you can afford to live further away in a cheaper location. And school kids can get to school in a car by themselves, so the total number of automotive miles is likely to go up.

Self driving electric cars mean that tunnels are more cost effective (no fumes, closer tolerances in tunnel sizing, greater capacity), so we can go 3D (no need for stop lights, we just have cloverleaf interchanges underground with sloped ramps so that passengers feel no lateral acceleration which means higher speeds are ok). Now the surface can be for pedestrians (like Venice) and of course no need for parking lots at all.

This means greater density is possible increasing the value of central real estate. But also means that further commutes are much faster, increasing the value of outlying real estate? Or does it, what does this suggest for real estate investing?

I agree that a narrower tunnel costs less (displaced rock is quadratic with tunnel diameter), but is it cheap enough to offset capital costs for all but a few extra worthy projects?

I'd love to have underground streets.... But $$$....

Look at the rate of subway construction in China[1]. Self driving cars + tunnel means the subway comes to you, the riders don't need to go to the station. In the US tunnels seem like a rare and expensive thing, but they're very commonplace.

[1] http://www.businessinsider.com/why-china-is-tunneling-a-mind...

The difficulty of public works projects in the US is a legitimate question. But it's a separate question from what self driving cars will make possible (especially for the 90%+ of people who live outside the US).

We did tunnels like you mention in Boston 15 years ago. The results have been great, but the costs, schedule, and construction hassles were a massive challenge to public consensus.

Tunnels unlike most structures can last for very long timeframes so city's that devote a small fixed portion of there budget could slowly build extensive underground networks on the 100-200+ year timeframe.

I have been thinking about this as well. One idea I've been wondering about:

* Will more remote vacation destinations, including owning a second home in these areas, be more in demand? Buying or renting a vacation house that's a 5-6 hour drive away may not be as ideal if you have to drive yourself there. If you own or can rent a self driving car though, the trip may not seem quite as far. Longer road trips may become more popular, and geographical far places may feel closer than they used to. Demand in coastal or mountainous locations currently far removed from cities may increase.

I think robocars will soon lead to robolandyachts. It'll be pretty tempting to rent a comfortable autonomous land vehicle that you tell the scheduling package that "I can leave after 1800 and I want to be at the Grand Canyon at 0800" - it tells you back the available vehicle systems and their amenities and what view the reserved parking spots at the endpoint have, and you and your family travel overnight without further attendance.

Are you just talking about RVs?

You can't go to sleep in an RV and wake up at a predefined destination.

So... you're talking about self-driving RV's right?

I think it gets worse before it gets better. Early generation self driving cars will move slower than human piloted cars, who take greater safety risks in the name of speed.

I live in outer London and commute into the centre most days by bus then train then tube. Most people around where I live do the same. I could drive, but it costs too much to park and the time is wasted time, whereas I can read on public transport. If I had a self-driving car, the downside of driving would be removed, so I'd be much more tempted to drive. I guess many others would too, but London wouldn't cope. The upshot of this is that London's congestion charge zone would inevitably be expanded and the charge to drive into London increased still further to keep congestion manageable.

I suspect that most cities that have good public transport would suffer similarly as commuters are tempted back onto the roads. The inevitable response would be London-style congestion charging becoming widespread, because few large European-style cities would cope with any significant fraction of the population switching from public transport to self-driving cars. In the end, due to such charging, I suspect we'd end up back in a similar status quo once the dust had settled, with most people still taking public transport.

On the other had, my kids would probably love to be able to get around the suburbs without me needing to taxi them.

Parking spaces could be tranformed to roads. And streets or even whole cities alloted to Robo Cars only and then used much more efficiently.

I grew up in an older, mixed suburb; a common complaint I've heard from young people who grew up in other circumstances is they grew up in huge multi-square mile monolithic subdivisions with nothing to do within walking distance as a kid. One interesting social effect of self driving cars would be "mom" (it always seems to be mom stuck with the job) no longer has to taxi drive their kid everywhere... Essentially every little kids has private limo service.

So if the closest public library is 5 miles down the interstate, a kid living there can get a similar life experience to my burb where the library was about 10 minute bike ride away. Wanna go to the park, fine. Wanna go to the library, fine. If kids still hang out at the mall, they can go there too. Friends house is 7 miles away? The self driving car won't care.

And a second-level social effect is the self driving car may eliminate helicopter parent attitudes toward kids being unsupervised... Well, maybe.

Another social effect people have not discussed is employers love sick employees to show up to work, cough on everyone to demonstrate their devotion to the cause, not to mention primate dominance (I can make you work when you're near dead, not vice versa). Often when I call in sick its not because I can't sit quietly and appear to be busy, similar to sitting on my couch, its because I literally can't safely drive when I'm this sick or medicated or sleepless or vomiting or whatever. Self driving cars mean workplaces are going to be even more like a sweatshop from centuries past at least WRT illness. The lower classes packed cheek to jowl sneezing and coughing on each other while moaning in pain... There are people who get off on other people suffering like that and they're going to really like self driving cars.

Some of the sensors involved cost $8000+ a pop [1]. The immense R&D expenses will need to be recouped from the vehicle price too. Expect sky high prices and little market penetration while the patents are in effect (20+ years).

[1] http://articles.sae.org/13899/

I think the impact of self-driving cars on real estate will vary wildly by geography as well as their adoption rate.

Urban roads in the US and Canada are quite wide and probably make it much easier to make an effective and safe transportation solution out of this technology. This isn't the case in Europe where urban roads are considerably narrower, regardless of city size. The same can be said about Japan as a whole and other large asian metro areas - Hong Kong, Singapore, Shanghai, Beijing, Seoul. Much tougher.

Europe and Japan also have a much denser rail networks than most of the US, which in turn will cause a different impact on real estate.

Driverless urban buses will probably be adopted much earlier as there are exclusive bus lanes in most medium and large metro areas throughout the world but I don't see it changing the dynamics in a fundamental way.

Scenario: Long-term, everyone switches to using robo-taxis and doesn't own a car. The robo-taxis can communicate and co-ordinate.

You're looking at an ~85% drop in the number of cars required (today only ~15% of cars are on the road at peak times). The cars are used heavily in the day, but still need to park somewhere at night. Either way, 85% less parking required. This can probably be fulfilled without street parking. Easier to have centralized charging without street parking.

With intelligent driving, you can probably get away with single-lane streets (each way), with either an expanded turning lane at the corners, or preferably round-abouts.

This leads to a lot more real estate available in cities. Something like 20% more, due to single streets and less parking. That's a big increase in quantity supplied, leading to a big price drop.

Since cars in motion take up strictly more room than cars at rest, you could accommodate all cars via dynamic street parking. At night, 6 lane thoroughfares could become 2 lane streets with the other 4 lanes devoted to parking.

Negligible impact on real estate.

If you live in the vast majority of the landmass that has no public transportation, your commute isn't long enough for a self-driving car to register on the factors in where you live. If you live in places with great public transportation, then it seems like it matters less--you now inherit issues (and cost) with parking and maintenance you didn't have before.

If you're looking for real estate impact, consider the aging population and the ability of self-driving cars to keep people in their houses longer than they normally would. Along with the factors above, I see the big beneficiaries being the service providers (anything from Uber to grocery stores to healthcare providers) who can now optimize location and distribution differently.

This might seem minor compared to the other real estate trends that self driving cars will bring about, but it's interesting to think about what might happen to all the land in urban and suburban American that is currently on-street parking. Hopefully we'll see some creative uses for it, such as gardens, bike lanes, or outdoor seating for cafes, instead of just letting it lie underutilized. Maybe on some streets we'll get something similar to the woonerfs [1] in the Netherlands, with streets being shared by cyclists, pedestrians, and slow moving self driving cars.

[1] https://en.wikipedia.org/wiki/Woonerf

The impact will be very wide, and certainly not limited to real estate, although there will be huge effects there to be sure. It's a transformation not less significant than the development of the Internet itself. Cars will not have to be heavily built to withstand collisions, we will not need controlled intersections, highways will be able to have fewer lanes, roadways and bridges will be able to be of lighter construction because of the reduction in vehicle weight, speeds will be able to be much higher than they are now, enabling fast travel between cities, etc. It's going to be huge.

> Parking lots in major cities are free'd up because they no longer make sense creating more housing real estate.

Buildings will continue to exist, this will only have an effect on large parking lots which will be torn down and replaced by housing, but these are already rare enough in cities.

> Traffic Congestion is reduced so commutes are shorter creating more demand for suburb living.

If you have to travel 30-50 km to the city centres, even autonomous cars will not solve the issue that this distance needs time for travel.

If self-driving cars make Uber or similar even less expensive, then people could just as easily prefer to live in the city without a car because it would be easier to get one on demand. Then they can have increased "walkability", which appears to be increasingly desirable.

On the other hand if transportation is more transparent then they may prefer the lower cost of suburbs and travel more.

There's different forces pulling in different directions, so it's hard to pick which way it will go.

Everyone seems to assume that robocars won't work in less densely populated areas, but I'm not so sure. In areas of little traffic it's possible to schedule trips further ahead of time. In addition, much greater speeds might allow a given number of robocars to serve a much bigger area. I'm sure there is a cutoff somewhere, but I think eventually it will be at a much lower population density than many expect.

I think what's going to change is parking lots. There will be less demand for parked cars. Also, eventually entire streets will have less parked cars, leading to nice-looking streets like there were before the 20th century. I consider all the cars akin to "garbage" since they are not being used.

What will really change real estate patterns is increased support for telecommuting, both with tools and cultural acceptance.

I've studied transportation network design in university. Here are some knowns:

1. You can, literally, not build enough highways for the demand. As you rack up lanes buildings get further away from each other. This means more drivers and more driving. Attempts to elevate or bury highways are only cost effective in very specific circumstances.

2. Peak flow (cars per minute) is at around 80km/hr, although sometimes increasing flow further is only a temporary salve, as exit points (the city downtown streets) block.

3. People are likely to respond to incentives. For example, dynamic pricing on the 407, higher rates during rush hour have pushed out enough traffic to maintain a high flow, even during rush hour. (Though some argue that this has resulted in more spillover to the 401, the closest competing route. I've yet to see a competent paper proving this either way.)

Consider that Uber already has ride sharing code in test markets. With algorithms that work together to minimize driving friction, and with surge pricing to deal with supply and demand at peak rush hour here is what I envision:

1. Way more capacity to get close to the city.

2. Much more expensive to get into the city, as capacity increases to the perimeter. With surge pricing, only the very rich will be able to afford a car alone. The poor will either get to the downtown core with a denser option (like a self driving bus) or they will get into the city at an earlier time, even if it means waiting around for work to start. Also, I see "surge insurance" being popular with the poor, since a 10x commute on a single bad day (due to weather or a unexpected road closure).

So where does this leave us? In the short term the core get's more expensive. The closure of a few parking lots is not going to make up for the built in surge pricing of self driving cars. When the price goes up, some companies move to the edge of the core. Leading prices to go up there as well.

The people that really get hosed are the people in the zone just outside of that. Suburban homes no longer command a large premium, since what would have been an unthinkable distance previous years (say, 175kms from core) is now completely reasonable. A software developer can sit alone in a self driving car that drives 150 km/h for an hour, get some work done, then pick up some co-passangers for the last 25kms to offset the cost. Get in work by 10:30am leave by 4:00pm, get some work done on the way home. Maybe watch a show. By 5:30 he's on his hobby farm.

Now all of this is in a relative context. If the Canadian housing market is too bubbly, and across the board it goes down then I'm just arguing it will go down more in the suburbs than elsewhere.

Until self-_fueling_ cars are a thing, all of your points, which are predicated on "never stop moving" are moot.

It's just a prototype but Telsa made an automatic refueling snake-thing that is super cool!

https://www.youtube.com/watch?v=ut3sELMOyTM

Nice find! Pretty much what I was looking for.

There may be a bit of overlap with self-driving ICE cars, though. I'm wondering if that would be more difficult to handle.

Either way, I had been looking for this story, ever since people started talking about modern self-driving cars. Don't know how it slipped through.

They have self-fueling gasoline cars in New Jersey.

I suspect that shopping centers will tax their member stores to pay for a self-driving bus that can pick people up in the neighborhood. They'll have to be at least close to wheelchair accessible so people can take carts on them rather than carrying their heavy groceries and home goods.

HN discussion from a little more than a year back. https://news.ycombinator.com/item?id=8059480

Reply on news.ycombinator.com