Saturday, November 27, 2010

110> Google's Robocars

First this note; I have recently been informed that this blog has twice triggered a virus alert for one reader. Has anyone other than this one reader had any malware alerts when visiting this site? I take this matter very seriously and have not been able to identify any malicious code. I have removed the “recent comments” feature from the sidebar, (again) since, in the Blogger forums, it has been suggested that such third-party “widgets” could be to blame, although no mention has been made of this (most popular) one specifically. Anyway, if you have received any alerts, please tell me about it, via email or the comments section. I am hoping it is a false alarm.

While I was out of touch last month, up at the cabin, an extraordinary thing happened that I just found out about. Apparently Google unveiled small fleet of driverless cars that had secretly logged 140 thousand miles of varied California driving with essentially no human assistance. It seems that the age of the Robocar is really here – at least as far as the science goes.

I am an unsuccessful contestant in a recent Google contest, which was about world-changing ideas. I can’t believe they picked the “Shweeb” concept over open-source, standards-based PRT. Now they have invested in Robocars. These people are no strangers to PRT, I just wish they would jump in with both feet. 

It seems to me that Google has “out-ULTra-ed” ULTra. After all, couldn’t Google’s modified Toyotas make the runs around Heathrow with ease? For that matter, it seems like they could just take you the rest of the way to your door. Back a few posts, I speculated that the real value proposition for a four-wheeled, pavement-driving designs like ULTra is really dualmode. Now it seems pretty clear that Google could clean their clock in that, and their present business too, if they wanted to.

In a way, my worst fears have come to pass. I believe that much (if not most) of the promise of PRT lies in the establishment of an alternative to roads, rather than making vehicles automatic. Regular roads are designed to carry huge loads, making them very expensive to elevate. Yet it is wildly inefficient to remain on the ground, where there are constant conflicts between people and cargo going different directions. There is also limited space. In most cities, well over 30% of the land is devoted to vehicles, either for roads or parking. This huge, paved landmass is an environmental nightmare, atmospherically, climatically, and hydrologically, not to mention a giant waste of very valuable real estate.

Because of the public’s resistance to having substantial overhead structures in front of their houses or businesses, it is essential for PRT track to have a minimalistic profile. This suggests something in the shape of a simple beam, and certainly not a wide, flat running surface, which would form an umbrella over the real estate below. (particularly at junctions)  Robocar makers may pay lip service to raised track, but practical realities say otherwise. Additionally, the removal of any way to “hook” into the track automatically makes travel in icy conditions something that just can’t be done at reasonable speeds.

With PRT being broadly defined to include four wheeled, steerable vehicles, advocates, (such as myself) are faced with the assumption, by many, that PRT is simply robocars on a designated roadway. This is defining PRT away from its original premise. Under the broadest definition the vehicles could even be gasoline powered. This is very, very far from the promise of PRT as envisioned by the early developers of such systems. No wonder people consider driverless cars as being an equally viable alternative to PRT. The PRT that they are familiar with has been stripped of most of its advantages. 

I just hope the good people at Google understand the unintended consequences that are inherent in their technology. Private driverless cars will encourage greater fuel consumption, because being free from driving will encourage other ways to pass the time, such as eating and drinking, watching TV, doing office work, etc. The resultant mobile office/living room will mean bigger, not smaller, vehicles. The comfortable and productive ride will encourage longer, not shorter commutes. It will encourage more pavement, not less.

Having robotic cars, especially with special lanes, may seem tantamount to a true PRT system to many, and this fact endangers PRT adoption. You can’t morph private cars on roads into PRT. PRT is meant to replace the need for more cars and roads. It is, at its core, efficient public transportation, which, in turn, encourages the building of truly efficient communities, leading to reduced environmental impact and greater prosperity.

Robocars will only help alleviate traffic, without cutting down on driving. It will make private car ownership even more desirable in emerging countries like India and China. This will eventually encourage even MORE driving and consequently MORE sprawl, just as faster highways have ended up doing across the U.S.  And mark my words, if they are privately owned, they will be BIG, powerful and very comfortable. I’ll take the camper!

11 comments:

qt said...

No virus warnings from me. Norton, at least, thinks you're good.

Now, comments:

I can’t believe they picked the “Shweeb” concept over open-source, standards-based PRT.
That's because you thought they were looking for real ideas. More than likely, they were looking for things that would give them a lot of hits from readers. Like the Popular Mechanics visions of the future.

In other words, you offered something practical and dull. Shweeb looked cool. No contest.

It seems to me that Google has “out-ULTra-ed” ULTra.
I agree. You could have added this, where we learn one of the reasons ULTra was scaled back at Masdar: because robocars had caught up, and Masdar didn't want to be tied down to an already-half-obsolete system that was essentially robo-golf-carts on dedicated sidewalks.

In a way, my worst fears have come to pass.
I agree again. I didn't quote more of the remaining post because I think you summed things up quite well, as far as you went.

And finally, my own thoughts:

I hate to say it, but Google is probably closer to the truth than either you or your readers really want to think about. They know what their "readership" thinks is desirable. And it's not what you want to sell them.

This is the source of my earlier, failed, attempt to get people thinking about where PRT could offer a kind of service the public would demand to have. I don't care how much better your mousetrap is. The world will NOT beat a path to your door unless their present mousetraps are completely unsatisfactory.

ULTra tried to sell the convenience of not having to drive. As you can see, the world's response was to build cars that don't have to be driven. So much for that argument.

qt said...

continued from previous comment
Even so, they've done better than any other PRT advocate/manufacturer. And that, I believe, is because everybody else starts with what they think is dreadfully important, and builds everything around it. Then they are shocked to discover that nobody much cares about their vision for the future.

So they try to persuade everyone that their vision is more important than any petty concerns that are merely popular. And when that doesn't work, they make some reluctant, partial compromise.

And when that doesn't work, they turn to some form of coercion.

This usually means trying to sell their idea to a government. But governments already have coercive transport systems. And the ones that don't, have ties to the people who build them. The PRT makers are basically trying to displace well-established institutions--and the representatives of those institutions, who are friends and campaign contributors.

In short, PRT, in its present stage of development, is a classic example of that old saw, "a solution in search of a problem." And until you find a problem PRT can solve--not in your own minds, but in the minds of the people who will BUILD it and the people who will USE it, you are wasting your time trying to find buyers for it.

Mind you, you may not be wasting your time trying to design the components for it. I'm all for what you're doing here, Dan--and especially the way you try to avoid tying your design to a particular "vision of the future." When--if--someone comes up with a use for PRT someone else will actually buy, we'll want a system ready to deliver the goods. And I have no idea what those goods will be.

I mean really. As late as the 70's, nobody really knew what a laser was good for...

Andrew F said...

There is one area where PRT (even ULTra) is a slam-dunk for, and that is for campus circulation in the place of APMs. APMs are usually oversized and expensive. PRT can do as well or better for a fraction of the capital.

So, PRT will at least have that market. The neat thing there is that someone will probably have the idea of extending that network just a bit further from the airport to the hotel or conference centre. It could snowball from there. But I agree that robocars will be a tough foe. And the idea of China and India building cities for a upwards of 600 million people in the next 20 years based on the car, even robocar paradigm is downright terrifying.

Supposedly there is a lot of interest in India. I guess we'll see how real it is in the coming years.

Bruce Attah said...

I don't think PRT and robocars are quite so much in opposition as you suggest. I think a system like ULTra would tend naturally to develop robocar abilities, to solve the "last mile" problem, and get people to the front door. At the same time, a robocar, especially an electric one, would tend to develop track-going abilities to save on infrastructure costs, to bypass other traffic, to maximize operating speed, and to increase range (power by pickup).

I don't think robocars would lead to an increase in private car ownership. Some of the first buyers of robocars will be of taxi fleet companies. The existence of robotaxis would tend to make car ownership pointless for most people.

In other words, robocars and ULTra-type PRT would tend to evolve into something resembling a hybrid of both, and by doing so, they would make both private cars and traditional mass transit obsolete, except in marginal applications.

This trend will be noticeable, I think, by 2020, and will be more or less complete by some date in the 2030s.

akauppi said...

Very good post, Dan.

I saw the pics a while back and after an initial "Wow, we're already there" I've even used the pics in my presentation lately. The point is, automated robocars cannot really move if roads are congested. And in most urbanising economies they are more and more so.

So they work "fine" in certain areas. Personally, I'd like one for a 160 km ride to my parents. They don't work fine in busy, local commutes with lot of pedestrians and/or manual drivers.

If robocars are provided as a *shared* service (like car clubbing, only yet smarter) they can be part of the solution, not the problem. Otherwise... at least they are not helping much.

Personally, I think the Robocar appoach makes sense in US. Infrastructure in most of the country is such that automated driving is a) rather easy and b) some form of driving is quite a necessity.

The rest of the world is a different story.
-asko


p.s. Andrew F: Where has India stated they are building cities for the cars? I think this is exactly what they're trying to not do. The transport mix in India is extraordinary (from cows to scooters to cars to lorries) and will continue to be so. Building for cars is something they shouldn't be doing - and I think they really won't be doing. Upgrading existing infra is another thing. Yes, maybe that will lead to what you're saying but I doubt they're actually *planning* for that. US should be a warning signal big enough - why repeat what seemed like a good idea in the 1900's? India is looking for new solutions which are green and yet provide wide coverage for its people.

China is indeed looking at robocars. Why not? They can be small, shared usage vehicles and one can upgrade them easier than PRT. These are challenges for the PRT industry to compete with.


Bruce: I agree.

akauppi said...

ps. I linked to this article from our website:
http://www.bmdesign.fi/en/links.html

cmfseattle said...

The batteries in 2getthere's Masdar vehicles need recycling every 3 months...

I hope robocars will sprout trolley poles, so they could utilize neighborhood energy storage systems. Buy your HDTV online, and get it delivered on a pallet by a robojack ;)

Dan said...

Dan the Blogger Responds -
A lot of really good insights! I am particularly struck by the notion of public or corporate robocars. One of the many problems with private, pavement-based transportation is parking, something that is vastly reduced if the same vehicle is used throughout the day for many patrons. It occurs to me that Google’s first step may be an automated taxi-service for their campus. I know they have been repeatedly approached about building a PRT system, but the PR associated with an automated limousine service would be every bit as good, and parking uses up their valuable real estate.

I guess my concerns are with the private variety. The migration to the suburbs is only kept in check by the expense and “down-time” associated with commuting. There are many downtown areas that are largely kept alive by a city’s traffic problems.

Batteries are, indeed, an issue, although a three-month life would indicate that they are not anywhere near state-of-the-art, or are way undersized. I agree with the sentiment of “sprouting trolley poles”. That’s one more reason why PRT (in my view) is a better solution.

“a solution in search of a problem…” I think the main obstacle is that the problems that PRT solves are themselves disconnected. No bureaucrat will make a dime more in pay if a city is more efficient. The people who plan roads do just that. Assessing new alternatives is not their job. PRT is environmentally superior, but those who appreciate that fact don’t tend to work with people who plan infrastructure. Who, in the chain, has incentive to reduce climate change? Or reduce the size of a Walmart parking lot? Or reduce the federal deficit? Or reduce the amount of mining and manufacturing that must be done to get people to work each day? The benefits are enormous when taken as a whole, but are miniscule to any of the key players, in the roles that they play, individually. There is even a perverse incentive to approve projects that involve greater costs, since the money comes in the form of politically rewarding gifts from out of town. On akauppi’s link (above) there is a paper by Nathan Koren, which describes the problem much better than I can.

I agree that, eventually, robocars lead to PRT. The electrification of the track, the faster, all-weather performance, the reduced cost of elevated track and the avoidance of ground based traffic, the reduced visual profile… It all makes PRT (track based) more obvious, while bringing the decision making process into the offices of the city planners. It worries me, though, that in the meantime being stuck in traffic will become no big deal, just an opportunity to get stuff done. That’s just the kind of mass lifestyle change we don’t need.

Dan said...

I might add that I’ll bet the cumulative time spent parking and walking to and from the car for Google employees is substantial, when you have that many people times that many work days…It might make a fair amount of economic, as well as PR sense for Google to equip the campuses with robocars. I wonder where the employees live? I am not familiar with the area… Also, I note that nobody has mentioned this malware warning thing…(no emails either…)

qt said...

Mr. Koren's paper was the starting point for my earlier "killer app" comment. It also influenced my comments here--that and some of the literature I've been reading on marketing and persuasion. One of the basics seems to be "it's hard to sell something to somebody who doesn't want it."

Another basic seems to be, "telling them they should want it isn't going to help."

Everyone here talks about persuading governments to build this stuff. Koren discusses some of the reasons that's hard. But you all talk about doing it that way for a very simple and painful reason.

Your neighbors don't want it.

Not "won't have it," just "don't want it."

They won't pay for it. They don't see why they should. They don't see anything in it for them. It's just another boondoggle--or else it's another attempt to take away their cars. And thus their options.

Some of them even see it as an attempt to herd them. To make it easier to control their movements, and thus their lives. I know a few of those myself--they see all transit initiatives in that light, whatever technology is used. And I'm afraid that (in the US, at least) a lot of the people who don't believe that still have a niggling suspicion that you see it as a benefit.

The only way you'll get those people on your side is to find a way to use PRT that they think is worth ditching the car for--at least part of the time. You have to come up with a reason they will support. Better yet--one so compelling that they will demand it--and it will be their idea to demand it!

I haven't seen anything so far that even comes close. I find it disturbingly apt that Google's idea of a great way to implement PRT is to turn it into a sport. And that Google's idea of a transport innovation worth really supporting has nothing to do with transit at all.

You're going to have to change that, somehow.

Andrew F said...

If people want LRT enough to build it, I don't know why the same desire for mobility can't support PRT. Koren made a point about the existing 'transit-industrial complex' conspiring against innovation. I'm not convinced. I do believe that most bureaucrats are conservative, and that transit unions instinctively oppose any form of automation (here in Toronto, we have automatic trains with drivers). You head both groups off by speaking to politicians. If they can deliver improved transport without taking away road lanes, and without any capital expenditure on the part of the government though Build-Operate-Transfer agreements, that is a pretty compelling proposition.