Hacker News Reader

I love this future because it completely routes around the public transportation dilemma we have in the US. The fact is that automated electric vehicles on paved roads would work perfectly here, could completely replace personal cars, would provide huge cost savings, not to mention saving thousands of lives a year from traffic deaths alone.

And this is something we could actually do. The government could take $X billion, get a cost-effective robotic car platform off the ground, start replacing vehicles on a 1-to-1 basis, and outlaw manual driving on public roads inside of 20 years. It would make existing electric vehicle technology perfectly suitable for trips of any length, since you can simply change cars at the charging station and be on your way inside of a minute. Completely feasible.

But alas... Inertia.

Automatic, cooperative routing also has great benefits over greedy time-minimising routing. In addition to the micro-level benefits (no stopping at traffic lights, fewer traffic waves) there are network-level benefits because the Nash equilibrium of user-optimal routing can be really crappy.

I highly recommend reading about Braess' Network Paradox for a great illustration of this.

http://en.wikipedia.org/wiki/Braess%27s_paradox

Well, the benefits of a private fleet of robotic cars is the ability to have many different providers - so I wouldn't want the government to have anything to do with it, save for setting safety standards for robotic vehicles.

There is no need to try and dictate the type of fuel, either.

But I'm interested to understand why people think a robotised taxi service would be much cheaper than human-driven taxi service. Is the driver really the highest marginal cost in the system? Does removing a driver really drive costs down that much? I wouldn't have thought so.

Oh, and your desire to ban manual driving is never going to fly. The most you would want is perhaps dedicated driverless roads where you aren't allowed to drive - much like bus lanes are now. That would allow for higher speed possible if computer systems can show they can safely transport occupants (or goods) at speeds higher than human drivers.

>But I'm interested to understand why people think a robotised taxi service would be much cheaper than human-driven taxi service.

The overall cost of the system would be cheaper because you don't need as many cars. A human taxi service has tons of taxis sitting around doing nothing for some amount of time, parked somewhere when not in use, etc. If taxi were just an automated service you would just need enough cars to cover peak usage and you'd have more cheaper options on parking (e.g. parking can be LIFO, no need for each car to have exit access at all times).

>Oh, and your desire to ban manual driving is never going to fly.

I disagree. Once automated cars start gaining traction it will be pretty easy to make commercials of crying mothers talking about how they wish their star football player son had only taken an automated car, coupled with real traffic statistics. In a 100 years people are going to find the fact that we used to manually drive cars insane.

Horses were more dangerous than cars, but you're still allowed to ride horses on the road. A total ban on manual driving is just not feasible in a democratic country. I agree that it's possible some roads would be automatic-only, but you're not going to get a ban on cars being driven by people.

You would need just as many robotic cars as you would current taxis. The demand shape isn't going to change just because you remove the driver. Yes, parking and other things will improve but it's going to be a small improvement, not an order-of-magnitude jump.

>Horses were more dangerous than cars, but you're still allowed to ride horses on the road.

They may have been at some point, but that hasn't been the case for some time.

>A total ban on manual driving is just not feasible in a democratic country.

I disagree. Maybe not this generation, but supporting manual driving makes the whole system less stable unless you make a road just for them, which is really expensive.

>You would need just as many robotic cars as you would current taxis.

I don't believe this is the case. I see taxis sitting empty for hours. In that time one taxi could have made several short trips. I'm certain that a computer calculating routes, etc., could manage to get the same work done as is being done now with less resources. That's why we made them, after all.

"Is the driver really the highest marginal cost in the system? Does removing a driver really drive costs down that much?"

Absolutely. In Austin, we have car2go, a car sharing program that allows one-way trips. Gas, insurance, maintenance, parking, and any other cost you could think of is included in the price. It costs about six dollars for me to drive downtown from my apartment. A taxi costs $15.

This isn't a perfect comparison since I walk to a car2go, but a self-driving car might need to drive to pick you up. It's pretty close, though.

Maybe the driver does add a significant cost from your example. I suspect this varies from city to city, as some cities have strict driver requirements (ie, London) while some others seem to just employ anyone with a pulse.

Another cost the car2go scheme is missing is the taxi licensing scheme, which most local governments use as a cash-cow by restricting supply and charging high recurring fees.

Add the fact that most robotic-trips are not going to be two-way and you're probably getting closer to $10 dollars for that trip, which, undoubtedly is still cheaper than with 'wetware' behind the wheel - but some people seem to be thinking this would be as cheap as mass transit, which to me is just wishful thinking.

If I can use China as an example, then in my experience, yes, the cost of the driver makes a huge difference. However, I am not entirely certain if the government is subsidising anything.

I think a lot of that will happen.

But what technologies power the vehicles or enable the roads are secondary. The efficiencies gained by reducing vehicle size will slash costs.

A great many people will commute in a single seat vehicle with 50 horsepower if its a quarter of the cost per mile. Especially if they aren't stuck with it the rest of the time. As long as they can just as easily get larger vehicles for other occasions.

As far as alternatives to conventional roads, we already have them in major cities. Some are elevated and some are underground. Today only mass transit uses them, but I fully expect robotic cabs to destroy mass transit. Even people who don't value convenience will ultimately go for them. So subways and elevated railways, once abandoned by trains, will be converted so they can be used by individual vehicles.

I disagree. See "one lane road" problem. Throughput will be very low and average speeds for urban areas will be close to that of bikes, at best, unless you're doing on/off ramps.

Also, re: many smaller cars... who will purchase/maintain all those? We don't have the infrastructure to support that on many levels.

Personally, I think a system like that in /I, Robot/ is much more likely.

People will operate businesses that own their own cabs or individuals their own personal vehicles. Just as we do now.

And the infrastructure we have now will continue to operate.

How will such an infrastructure be built and be operated. Its a mystery. But its already been done by private enterprise.

As to why only one slow lane could be used, I don't see why that would be. Perhaps you could explain. We have many lanes operating now and robotic vehicles would simply use the same lanes.

Perhaps PRT has these limitations. Robotic vehicles that operate like the vehicles that operate like the vehicles we already have wouldn't.

Of course, we do not yet have robotic vehicles that can be trusted to operate in rain or other extreme conditions and that can discern their environment well enough to be safe.

But once we, there's no reason to follow the PRT pattern. We'll simply follow the patterns we already have. Only with much more sharing (cabs) and much more efficiency.

This is a great point that I haven't heard yet. Reducing the cost by a factor of 4 might be a bit to optimistic though.

It may very well be overly optimistic.

But many features of today's automobile are there because of human needs for performance. Optimize for efficiency for a moped-like 3 wheeled vehicle and you may get something completely different.

  http://www.niutoday.info/2011/09/14/supermileage-team-proves-nation’s-best-again/

I'm not saying cars will actually be like this, but robotic vehicles won't need mirrors or more than one seat or even a steering wheel. A lot of things are could change.

Absolutely. And your comment doesn't even mention the tens of thousands of lives that could be saved yearly, nor the millions of acres of paved wasteland that could be re-claimed for use by humans and real human communities.

I hope someday an america jammed to the gills with empty, idle hunks of steel looks as bizarre and wasteful as it actual is.

Surely that material can be recycled?

Once you switch over to autonomous taxis, so many possibilities open up:

Imagine eliminating most visible parking. A small fraction of the world's vehicle fleet is in use at any given time; the rest of the time it's taking up a parking spot. Remove that and we can have wide sidewalks and bike lanes, parks and new development. Suburbanites get an extra room in their house and a bigger yard.

Buses and light rail become obsolete, but heavy rail transit becomes much more practical, because the door-to-door problem is solved. A virtuous cycle of improved convenience, leading to greater ridership, leading to more trains and shorter headways could result. It makes a lot more sense to ring the bay with a 150 mph train when getting to and from the train is easy. (Then again, it might be easier just to run the taxis at 150 mph on dedicated roads.)

You could get rid of passenger trains and just have the cars drive right in when they chose a train route.

Also, for parking, we can take advantage of LIFO. Today, every single car has to be able to get out independent of the others. With robot cars only the last one of a given type (e.g. van, pickup, sedan).

I love hunky-doodle blue-sky plans too - they don't have to deal with reality.

In reality, most people leave for work at the same time. So they all need a car at the same time. So we need the same number of cars as we have now.

Carpool? Did I hear somebody say carpool? With robotic drivers that works ... exactly as badly as it works now. My three nearest neighbors work in different TOWNS, not only different buildings or neighborhoods.

So I jump on Google Taxi, log in, reserve a car to show up and ... it says "next available taxi arrival, 2 hours before or after desired time". So I either get up at ass-oclock in the morning, or am chronically late for work. </rant>

You can do all of this right now either using a shared vehicle service (you drive) or getting a car + driver (whether limo or taxi). I'm curious as to why people think that robotising a car would lower the cost of hiring dramatically to the point where you were using it to get to work?

It's my understanding that a taxi driver isn't the major cost in a taxi fare, and that fuel, capital cost + licence costs are the principal component in a taxi fare.

I doubt this sort of thing would make much of a dent in car sales, at least that wasn't offset by sales of the automated cars.

The biggest category for sales of cars is for people who regularly feature a combination of long & short drives in their life. This is typically not your average city-dweller, despite having larger numbers of people within cities. The average mum+2 kids isn't going to trade the family car in any time soon for this type of service, not when the cost of keeping and running an average family car isn't really that burdensome.

A 3 mile cab fare in Berkeley is $20+. The all-in cost of driving a basic private car is around 55 cents a mile (or 90 cents for a BMW). That only leaves license cost, so unless you need a $500,000 medallion, the rest goes to the driver. (And as long as the driver works for himself, the medallion is a capital investment which he can sell later.)

But the cost of taxi licences is significant in many cities. I doubt any local/state/federal government is going to let the possibility of driverless taxis proliferate without similar licensing schemes.

I agree that having a car running privately is a lower per-mile cost, but that discounts the number of 'dead' miles that a for-hire car drives - the private car always takes the person where they want to go, while ordering one up via the internet is going to at a minimum double the per-mile cost, because only half the miles (as an average) will be utilised. The car will also not always be in use because demand will never be that smooth, so the in-demand times have to be priced higher to make up for the quiet times. Then there has to be a higher insurance cost, plus a margin to cater for damage to the cab, booking systems, marketing, etc etc.

It's not that hard to see where a cost goes from the 50c/mile to the ~$4/mile which isn't far under the taxi cost.

Taxi drivers make just over minimum wage in most places. While the taxi (plate) owner does make a decent return if run well, it's not like there are a pile of taxi billionaires around. I'm just not convinced removing the driver is going to change the cost by an order of magnitude.

Well, like you said, the fare has to make up for the time the driver is idle, so that problem goes away. Licensing is done to ensure passenger safety, and sometimes to restrict supply. The computer is not going to drive drunk or kidnap you, and excess unproductive machinery can just drive itself to a city where there is more demand.

And the dead miles problem is not a big factor in places like Manhattan where utilization of taxis is high. If this system were pervasive, you could simply maintain a car or two per block.

The biggest issue would be dealing with the peak demand. Fortunately, heavy rail is complementary here, and it's only marginally less convenient when you have quick on-demand service to and from the train station. Demand-based taxi pricing would sort things out.

Automated taxis will be cheaper and more convenient than owning a personal vehicle. The taxi driver is a significant cost.

If you're sharing a car, you're paying a smaller share of the capital costs. Most people don't do that today because it's inconvenient. If cars drove themselves, it'd be convenient.

> I mean hell, I know I wouldn't want a car if I could just have a subscription to a car. Especially if it wasn't just like Sedans. Imagine being able to rent a truck immediately at a moment's notice, then ride back home in a Sedan, all without having to worry about driving anywhere?

It's called Zipcar or your local equivalent.

Zipcar isn't prevalent in the US outside of major cities. But I've used it and it is similar to what I'm talking about, just imagine that but fully automated and run by fewer hipsters.

The definition of hipster has inverted if grey haired guys with MBAs are now hipsters.

> Zipcar isn't prevalent in the US outside of major cities.

Neither will any robotaxi or robocar system initially. I suppose some will view being run by nerds as an improvement over being run by hipsters.

And Google could probably provide the service for dirt cheap or even free by charging commercial venues per person delivered (at least if your business wants to be at the top of the list for "sushi.")

More likely is that hot meals will be delivered by a robotic transportation system. The same thing thats happened to a night out at the movies will happen to eating out at a restaurant. Except instead of a disc being delivered by snail mail, tiny robotic vehicles will text you when your meal arrives at the front door.

Or we could stop wanting to live out in the boonies and actually build things that people can walk to, then stop being so lazy.

I would prefer to be able to walk to everything I need. But unfortunately, that usually comes with a lot of other problems I'm not interested in.

The fact is that in a modern economy many things are shipped from business to business and from business to individuals. Laziness has nothing to do with it. Everything thats available in local businesses also has to come from somewhere.

Many things must go through many hands before they become available across the street. There's no reason a human being has to chauffeur things from place to place.

Or you'll watch a couple of ads during the trip ;)

Only if there's an option to pay for that not to happen.

http://en.wikipedia.org/wiki/Personal_rapid_transit

Car-as-a-service. Nice.