This past weekend, the New York Times ran a story that contains a harbinger of things to come. The story, titled Google Cars Drive Themselves, In Traffic, describes how a team of 15 Google engineers are actively pursuing technologies that will allow cars to drive themselves.
And this isn’t just fanciful fiction. The team has already been hugely successful. Without human aid, seven test cars have already driven 1,000 miles within mixed traffic and another 140,000 miles with only limited human intervention.
The benefits are obvious: Less traffic, higher speeds, fewer accidents, reduced costs. This would be actual Personal Rapid Transit, not the fictitious one dreamed of by the PRT community.
The researchers are cautiously optimistic but realistic; “even the most optimistic predictions put the deployment of the technology more than eight years away.”
Eight years is not, however, a lot of time. Eight years – give or take – is the time it takes a city to build one LRT line. Even double or triple that number and you’re still talking about a span of time less than a generation long.
The idea of a driverless car has been around for decades, but was always just that; an idea. For most of the idea’s life, the technology simply didn’t exist to transform it from dream to reality. Furthermore, driverless car technology isn’t one technology, it’s a package of technologies.
For driverless cars to work you need cloud computing; fast, wireless and cheap mobile communication; minute-by-minute traffic data; geographic information systems; low cost optical sensors; and a whole host of other things that are far more complicated for you or I to ever understand.
So if there’s one company equipped to undertake this Herculean challenge, it’s Google. Remember: Google isn’t just in the search business. Google has Google Maps, Google Earth, Google Traffic, Google Streetview and the Android mobile operating system. And Google’s only 12 years old.
Just barely into its second decade of life, Google has utterly transformed how we live. True, most of those changes have occurred in virtual realms, but the company’s influence on the actual world is becoming more and more apparent. As the world’s most efficient (and profitable) collector of data, it would be smart money to believe Google’s the one to make the driverless car a reality.
So what then does this mean for public transit? Likely it means the end – at least as we currently know and understand it.
Taxi companies would be the first to adopt the technology. They’d have the fleet of vehicles, dispatch facilities and communications technology to roll-out the technology quicker than anyone else. They’d also have the motivation. Drivers are costly.
Taxi drivers wouldn’t disappear over night. Legal requirements would need vehicles to be supervised by humans, but as the technology advanced and scaled, those legal requirements would quickly vanish. The taxi driver would go the way of the travel agent; it would simply cease to exist.
At that point, all bets would be off. Public Transit would still be caught in its current bureaucratic nightmare. Fares would continue to increase; infrastructure costs would continue to skyrocket; and unionized drivers would continue to maintain a stranglehold on any ability to control expenses. Politics would continue to prioritize routes with little actual value except for political gamesmanship and it would still take a decade or more to build even the most modest fixed link transit line.
Without the expense of drivers, the private taxi industry would be in a position to slash fares in an attempt to capture the lucrative public transit market. Tax payers, meanwhile, would find themselves questioning why they are subsidizing the Public Transit industry when the private taxi market serves them so much better. And that would be the end of it. Public Transit ridership would plummet and subsidies cut. Public Transit would become nothing more than a 100 year old blip in the history of humanity.
Will this ever happen? Who knows, I certainly don’t.
But it’s worth considering how poor a competitor Public Transit is. Public Transit, it seems, refuses to acknowledge the need to compete. It’s amazing that an industry so subsidized by government and so in a position of monopoly cannot currently provide better service. That’s not a condemnation of Public Transit, but it is a criticism.
Public Transit has to understand that while they’ve been arguing about bus versus train for the last 20 years, the world has changed to a point where it’s now possible for a car to drive itself! Meanwhile, the Personal Rapid Transit industry still can’t get 20 automated cars at Heathrow to work.
The future of Public Transit isn’t lost. The industry will still have the developing world to ply their trade in and there’s likely always to be a place for high-capacity, high speed transit systems in urban areas. That place, however, will become severely limited and specialized.
Public Transit is in desperate need of a reinvention. If the industry doesn’t respond to the clear challenge before them, you might one day open your inbox to find an invitation to join something called Google Transit – Beta.
20 Comments
“It’s amazing that an industry so subsidized by government and so in a position of monopoly cannot currently provide better service.”
Have you considered that private cars are subsidised by the fact that governments build the road system out of general taxation? Those who don’t drive (or who drive little) subsidise those who drive a lot.
It’s little wonder that transit has trouble competing when government refuses to extend the same subsidy to it that it extends to private car use.
Adrian,
That’s a fair point. But the point is that transit is massively subsidized just as cars are. Despite that massive subsidy, transit is still losing this battle to the private automobile. Thats a huge problem.
http://www.spiegel.de/video/video-1088474.html
http://www.youtube.com/watch?v=xb7G07ivPZ8
For the interested: those are videos about a driverless car. A few days earlier in Braunschweig a car named Leonie had “her” tests and they all went fine.
By the way: I can see parallels between those videos and the one Steven posted of Disneys vision of the future.
“Despite that massive subsidy, transit is still losing this battle to the private automobile. Thats a huge problem.”
Depends where you live, right? Perfect example is Dubai.
When the city was booming everything was done by cars. So people could pick a place somewhere in the desert and build their houses there. They didn’t even have infrastructure for water or internet there.
When Dubai decided to go green around 3 years ago it all changed.
Dubai Marina for instance got remodelled (although only around two years old) and they used district cooling and infrastructure for water etc.
This area became very popular, because it was working.
Last year the Dubai metro opened and connected over the main vain, the Sheikh Zayed Road all the important areas of Dubai with each other.
Dubai Marina finally became one of the best working areas. Next to it Media City is situated. Every mayor district has an own stop. That is the way how public transit is working. It all depends on the accessibility – and in relation with it the times of travel too.
LX,
I’ve heard from a colleague who lives in Dubai that the metro is suffering from severe lack of ridership. He says the reason is due to the long distance between stations coupled with the extreme heat. It makes walking very unpleasant.
But I will admit to having no confirmation of this, its merely a story I’ve heard.
That’s right. Not all stations are perfect right now. The one in Marina is really good and another one connected to the Mall of the Emirates too. They even got cooled-pedestrian bridges.
My point: if dwelling is directly connected to those stations it absolutely makes sense and is working fine. My guesses are later on the small buildings around those stations will be replaced by good working and connected high rise building-communities – so in the end an agglomeration of buildings (like satellites) will make this city working.
But let’s focus on the actual idea behind this concept and not the more than questionable idea of living in a desert.
I think it would be interesting to know how far/long people would accept to walk to get to a well connected point for further transportation. Point being: when do you really need a PRT?
I can see advantages in driverless cars, but it’s just another way of driving a car.
Even if it all will be managed by one unfailing controller for the infrastructure system – I see people losing flexibility (a good example is the transportation vision in the movie ‘I Robot’) and am not sure if we all like that. On the other side: if this system is really working, there would be no traffic jams, no accidents and very calculable times – which of course would be really good.
i’ve always imagined cars that can drive them selves as needing less road infrastructure, specifically stop lights… similar to how people move through crowds, the cars would be able to communicate and “look at” each other in a way that means everyone can go through an intersection at the same time.
an easy comparison would be to equate this to the “scramble” intersections in toronto (http://vimeo.com/1626058).
while this could make driving much quicker, but i see it posing a problem for pedestrians. maybe it would mean people could cross anywhere and anytime, or maybe it would mean pedestrians would also have to be “linked in” to this car communication system.
If Google (with the help of Stanford and CMU researchers) can devise technology package to operate a motor vehicle in mixed traffic, with signs and signals, pedestrians, cyclists, at varying speeds and conditions; I imagine a future where public transit (taxis, vans, shuttles, buses, LRT, metros and commuter and intercity rail) are all guided by similar systems adapted for their particular conditions. In fact, I can easily see the rails being the first beneficiary of this technology. Its inevitable. Now all we need is to have the Japanese perfect their humid robotics project and we can say so long to any human intervention in our transit experience.
Nice article.
I wouldn’t quite declare public transit yet.
Even if cars would start being driverless in a decade, it will be a long time before people are not allowed on the roads anymore.
Also, new transportation technology has in the past often helped transit as well. That is, even a disruptive technology (i.e. the car) has not destroyed transit or made it obsolete, but had the potential to actually improve it. So driverless cars means driverless busses, which sound more likely given that they would drive on a subset of the street grid. Thrown in with some PRT ideas but no necessity to have its own right of way could actually be pretty useful.
Interesting scenario.
However there are issues such as congestion (no matter how automatic cars are, there is a limit to roadway space) and issues such as peak oil. Even if they were electric cars, it still doesn’t seem like a wise use of resources (space and energy). Do we really need a large chunk of metal and plastic to transport a person?
This could also provide a tremendous benefit and increase the competitiveness of public transit. This would significantly reduce operating cost allowing transit to be run with greater frequency and span of service hours. And given the amount of mass that has to be moved to carry a person, I’m sure that given a certain amount of ridership, a transit vehicle would use less energy resources than a single person riding in their own vehicle. Since whether you travel alone or not you aren’t paying for a driver, transit should be competitive price-wise.
Also, if public transit agencies work with these taxi companies to provide an integrated fare system for the entire region, these automatic taxis could replace under-performing transit (I’m thinking bus here) routes in suburban areas of the city that don’t have many riders. Then resources could be dedicated to providing service in areas of the city where the amount of traffic and density merits public transit that run with sufficient frequency and span (of service hours) to be competitive against the automobile. And if this corridors received enough use, you could justify giving them their own ROW giving public transit the speed advantage as well.
So rather than being the end of public transit, I think driveless technology could actually significantly benefit public transit.
I’m also in the camp who thinks this technology could be good for public transit, at least broadly conceived. Other people above have touched on various relevant issues, but I would particularly emphasize the potentially disruptive nature of any technology which frees households from the need to own and store their own cars in order to have car-like functionality.
Owning and storing your own car has large fixed costs in relation to the marginal costs of use (although of course that ratio depends on things like fuel prices and such). Accordingly, if you have to own and store your own car for some purposes, you will tend to then end up using that car for a lot of other purposes as well–indeed, you might well implicitly organize your entire current and future-planned lifestyle around the assumption that you will always own and store your own cars.
But if you can have car-like functionality without owning and storing your own car, and if the entity providing this service can spread its own fixed costs through to the marginal units of service, then suddenly all this may change, and your best strategy may end up being to plan to use this car-like service if, but only if, it is the most efficient mode for a particular purpose from a systemic perspective. And you might well be willing to organize your entire planned lifestyle around the assumption you WON’T be owning and storing your own cars, although you will be planning to use this car-like service when it is the most efficient mode.
And in such a world, there may be many more opportunities for public transit (even defining this car-like service out of the realm of public transit) to provide valuable services. In fact, this car-like service may be less of a competitor for public transit, and more of a complement to public transit, such that we may well expect increased market penetration of this car-like service to correlate with increased, not decreased, demand for public transit services.
Of course if you assume that public transit authorities will be completely incapable of responding appropriately to such opportunities in a timely fashion–well, then you are basically assuming the conclusion (that this sort of technological disruption will destroy public transit). Personally, I think it is possible, maybe even likely, that a confluence of factors will cause a radical reorganization of how we fund and operate public transit, such that public transit authorities will in fact ultimately become more nimble in terms of seizing this sort of opportunity to prove a complementary service.
By the way, some of this same logic applies to current wireless-enabled carsharing services like Zipcar (basically, the vision here is of something like a carsharing system where you don’t need to be particularly near a storage location in order to use the service, since the closest available car can drive to you from wherever it is being stored). And I believe Zipcar has in fact found their service to be complementary with public transit.
It’s interesting how much reaction this post has received. I think everything that everyone is saying here is quite thoughtful.
I think the strange thing is how unadvanced our public transit technologies actually are. Sure there’s signal priority schemes (and their dubious track record) and automated ticketing, but the technologies and vehicles themselves have evolved very little.
Meanwhile, every other non public transit technology seems to be advancing at an astronomical rate.
Private taxi service merging with public transit… quite a likely possibility.
“Private taxi service merging with public transit… quite a likely possibility.”
Seems like I’ve heard about cases in Germany where this already happens. Such as taxis running bus routes late at night when it would be unprofitable to run a bus and accepting the normal bus fare or something like that.
Just an addendum, but I’ve noticed a bunch of other commentators in various places making the connection between driverless cars and carsharing.
Steven Public Transit evolved a lot.
While there are some prototypes of automatic cars, fully automated metros carry hundreds of thousands passengers every day. Major airports rely on fully automated transit systems. The power electronics, the motors everything evolved. Low floor buses and trams are standard now. Modern transit vehicles and station are fully accessible for handicapped people. Information systems are much better than there where a decade ago. etc. And since over hundred years urban rails transit is electrified with the possibility of running with environmental friendly energy.
And if you want you can have very fancy technologies like maglev. You just need to order and build them. I also think a autopilot for cars would be very nice but i don’t think any authority will approve them. And if approved they need the same or more redundant electronic like automated metros or aeroplanes, which means nobody could afford this kind of car.
I still don’t think it’s a sure bet this technology will even get off the ground. Does it work? Certainly; that’s been proven. But will people accept it? A much better question.
Ultimately it’s very simple, would you cross the street in front of a driverless car? Come on really picture it in your mind. Better yet let’s make it a twenty year old driverless chevy rust box with a huge dent in the fender and an ominous rattling coming from the engine. There’s no one in the drivers seat. In fact there’s no one in it at all. Feeling confident?
And just so we’re all clear, these things are going to kill people. It’s inevitable. No technology is perfect, and these will be handed to people who won’t take care of them properly. They’ll still be on the road twenty years and five irresponsible owners later. Once in a while a driverless car will go wrong and kill someone. Of course drivers do that too. They do it a lot and logically driverless cars will be far safer even if they sometimes drive through somebody’s house. Still, would people accept that a five year old girl was just run over by a computer? That’s a tough one.
Erik,
All valid points. How this technology is ultimately “rolled out” will determine its success or failure.