Comment by aetherson | Hacker News Reader

Comment by aetherson | Hacker News Reader

Parent Full thread

I don't think that "driverless" segways have much of a niche in most versions of a driverless future.

The potential for segways or at least segway-like-transportation-vehicles are, it seems to me, in a rail-dominated world. The fundamental problem with rail infrastructure is that rail only takes you to stations, and that stations are expensive (along several axes) but since you can't take your car on the train, you're limited to a very small radius around the station. Segways would extend that radius dramatically, making a functional rail infrastructure de facto cheaper.

Most versions of "driverless cars" futures destroy rail-based infrastructure. They take all the advantages of rail and give them to the (already more popular) cars. At that point, your car can drop you off right next to your destination (and then go park itself elsewhere if necessary), so why do you need a segway?

The versions of "driverless cars" that don't destroy rail-based infrastructure, I think, make "driverless segways" unlikely as well.

See my comment above, but the idea would be that you build your architecture, from the building to the town/city level, around the more efficient segway-sized form factor as opposed to the bulkier car-sized form factor that's better suited for distance travel.

Predictions based on, "First, we'll completely redo our architecture from the ground up" are likely false.

It seems like you have a (pleasant!) vision in mind, but not like you've taken a lot of time with the practicalities of it or how we get from here to there.

So, first: the size of cars isn't principally about distance travel. Cars are built for speed (as soon as you're doing about 20mph, wind in your face becomes annoying for most people most of the time, and noise becomes high, and safety becomes an issue), for protection from the weather, and for cargo storage. Driverless cars may well become a little smaller, but there are plenty of reasons they'll stay fundamentally cars, not segways.

Second: you can't build your architecture from the building to the city around segways. We already have cities. We can change them incrementally, but redoing them from scratch is just too expensive (again, on a number of axes) to contemplate. And incremental changes aren't going to go your way. People will get the driverless cars that work with how the cities work TODAY, and then those people will become a constituency for incremental changes to the city that will make the city MORE car-oriented, not less.

Seriously, the full-featured, relatively cheap driverless car future is one in which public transit declines drastically in importance. It is absolutely not one in which people restructure their cities around transit.

Speed, weather, and cargo storage are mostly about transportation over...distances. For most shorter hops you need none of those things. Hence my statement.

Of course existing cities wouldn't be rebuilt from scratch. But cities change all the time.

Driverless cars certainly are better suited for the way cities are today. But as it becomes obvious that there are alternatives that don't require as much wasted real estate as cars use, and that crucially don't give up the personal freedom that cars provide, cities can and will adapt to that.

The city of 2035 will likely be moving down that kind of path, whatever the exact details look like. New development will be designed for it from the ground up.

This is just how transformation happens.

Speed (at the ranges that we are talking about) is important at the distances that we are talking about. Weather and cargo storage are totally distance-independent.

But let's take a step back. We're imagining the sort of default rosy driverless-vehicle future, right? Driverless cars can be made that are fully autonomous, at least as safe as human drivers in essentially all conditions, and not terrifyingly expensive.

We agree that in that case the driverless car is strictly more attractive to riders than the drivered-cars of today are, right? So if the value that a person today puts on having a car is A, and the value that future person puts on having a future car is B, then B > A? Indeed, B is much greater than A?

And we also agree that today, cities are pretty car-oriented? So A is large ENOUGH that cities pander to it to a large degree? And there's a positive feedback cycle here where when cities pander to cars, that makes cars actually more valuable compared to the competition.

Now what you're saying is that when we replace A with B, where B >>> A, cities will (eventually) pander to cars less. Why?

The thing that you suggest is that cars waste real estate. Which, certainly, they do. I do note in passing that driverless cars might waste less real estate (for example, they could be sent to more distant and more inconvenient parking structures, or the utilization of people / car could go up with a rides-for-hire model). But whatever, at the very least the real estate efficiency of driverless cars is no worse than the real estate efficiency of regular cars.

So meanwhile, today, the rail + maybe something like segway model has value X. We agree that X < A, right? That in America at least cars provide more value than transit, and so people get cars, and so cities pander to cars. Now you seem to be envisioning that in the driverless future, rail + segways have value Y, and Y is not only > X, it's also > A, and in fact > B. Why? What is it about driverless segways that is so great? Unlike driverless cars, they don't seem like they provide much value for their users (you aren't going to read a book while your driverless segway takes you to your destination, and though maybe they'd be a little better when drunk).

And, further, we're saying that driverless segways are that awesome even if they have to reverse some inertia of cities currently being designed kind of "against" them, and indeed, even if they have to reverse yet more inertia because the FIRST thing that happens in our driverless future is people buy driverless cars and become a constituency for even more city-pandering-to-cars.

I don't see it. Maybe if it takes a long time to get to driverless cars, because right now I think that the car domination of the world is getting a bit milder. Maybe if the current trend continues for a few decades, and then cities have become more hostile to cars, driverless or otherwise, the increased value of cars won't be enough that it will get over the inertia.

But if we get good driverless cars in the next decade or fifteen years? In-city rail is just going to die in almost all cities. And once it's dead, it will be dead for a long, long time.

You're kinda taking the "faster horses" side of things, if you know what I mean.

You're comparing the real estate efficiency of cars to driverless cars, which is not what I'm doing. I'm comparing the RE efficiency of cars to segways, and segways win that hands-down, no comparison.

The other benefit is psychological. You transform a development from being centered around cars to being centered around people. Even a person on a segway is a person. A person in a car is a car. This, I would maintain, is the larger benefit to average citizen. It's a more desirable experience.

Developers win with more usable real estate, citizens win with a much more pleasant scale and experience.

The segways would be for intra-area transportation, the driverless cars for inter-area.

The bigger assumption you're making is that people will "have" cars, and I'm not sure that will be the case. For me, that's the biggest unknown about transportation in the next couple of decades. Will people own cars, or will they merely use cars from a large pool of cars (either civic or corporately owned)?

Uber, for example, is starting to promote the idea of not owning a car at all to young people. They've been doing a good job of hyping the convenience of that lifestyle, and a limited set of people have bought into it already. I also recently noticed they're starting to play a prestige angle with black SUVs and so on. They're actively playing that up, which fills the other need personal cars currently serve.

Uber's endgame is very clearly a fleet of autonomous cars, and some cities are making very early sounds about building their own fleets.

I don't know how that's going to shake out. There are powerful forces pushing both ways.

But, if personal car ownership does become a thing of the past, a lot of the pressure you imagine for city development to have an even greater universal (as opposed to limited areas) accommodation for cars goes away.

The other major unknown is if cities simply become less desirable in 20-30 years. Given increasing automation, the nature of work may change to a degree that some of the economic incentives for city living decrease. That trend, if it even builds at all, will be slower to develop, but potentially even more powerful.

You're misunderstanding my comparison.

I'm saying, "real segways are more space efficient than real cars. Despite that, people prefer cars to segways (by a gigantic margin)." So from there, I get to, "future segways will be more space efficient than future cars, but I don't see that turning people towards future segways -- because we can already tell that people don't value space highly enough to make them want segways, and it's going to be pretty much the same comparison in the future."

Here's another bit of pseudo math: real cars >>> real segways. Driverless cars >>> real cars. Driverless segways ~= real segways. Thus driverless cars >>>>> driverless segways.

I'm not taking a stance on the ownership model of cars, either. I don't know if in a driverless car world people would own them or rent them for hours or days at a time or rent them for the duration of a single ride. I'm just saying that people will USE them a lot. And regardless of who owns the car, if everyone USES the driverless cars a lot, then cities will be built around cars.

And cities that are built around cars are hostile to segways, driverless or otherwise.

You say this intriguing thing: "The segways would be for intra-area transportation, the driverless cars for inter-area."

I think that my fundamental point is that that doesn't seem true. If driverless cars are highly available to people, I think they'll just use the driverless cars end-to-end. Your driverless car trip will put you at the doorstep of wherever you're going. At that point you don't need a segway.

Segways are for when your inter-area transportation CAN'T get you the last mile of where you're going -- that is, they're for rail. And rail, I think, is dead if we get driverless cars soon.

> You say this intriguing thing: "The segways would be for intra-area transportation, the driverless cars for inter-area."

Yeah, this is the root of our disagreement. As I see it, the reason this doesn't happen now is a matter of convenience. As things stand now, there's the friction of transitioning from car to segway. In a fully automated world, I see that friction largely disappearing. You get out of the car, it goes away. The segway is waiting for you. It delivers you and goes away itself. There's very little friction.

It's true this only makes sense in certain kinds of living situations. Suburbs seem unlikely to go this way (though there was a brief trend last decade to more pedestrian-y suburban town center development). But for urban areas, this kind of development increases what people like about urban living, and decreases what they don't like. I don't think people know they want it, but I also don't think it's something they've been given much to consider. When they have, like 3rd St. Promenade in Santa Monica, it's been hugely popular. Segways would allow you to scale that up considerably, and expand it beyond retail and entertainment, possibly to an entire city center.

We're in agreement about the future of passenger rail. It looks bleak decades down the line. Interestingly, Mayor Garcetti of Los Angeles said much the same just a few months ago. He wondered in public if it made much sense to keep building massive rail projects given that automated cars were coming.

http://laist.com/2014/09/30/mayor_garcetti_wants_to_turn_wes... "How do you spend billions of dollars on fixed rail when we might not own cars in this city in a decade or a decade and a half?"

Maybe for legacy applications where driverless cars can't fit. They didn't tear down and architect most of Europe to make room for cars so why should driverless cars be any different? It would basically be the Vespa of the driverless age.

Reply on news.ycombinator.com