Clinging to Fantasies to Avoid Admitting the 20th Century Is Over

Sep 4th, 2012 | Posted by

Update: Brad Templeton responded late Friday afternoon in a comment on this post.

Today Forbes’s website published a rather odd article by Brad Templeton of something called “Singularity University,” a research outfit in Silicon Valley that claims to be working on the technologies that will solve human problems. The article claims that high speed rail is outmoded technology and will be supplanted by things he claims are futuristic but are, in fact, even more archaic than the bullet trains he derides. What’s he’s actually doing is desperately finding ways to prolong a failed model of transportation and land use, even as the public is moving away from sprawl and driving, unwilling to admit that electric trains are where the future lies.

It's true, there were no humans inside

The reality of the world, however, is that technology is changing very fast, and in some fields like computing at an exponential rate. Transportation has not been used to such rapid rates of change, but that protection is about to end. HSR planners are comparing their systems to other 20th century systems and not planning for what 2030 will actually hold.

Actually, we are planning for what 2030 will actually hold. It’s these supposed visionaries at Singularity University who are so busy looking backward they’re not paying attention to what’s actually happening or where the trendlines are pointing. And as you read his article closely, his biases – in favor of sprawl and utterly ignorant of rising demand for urban living – are obvious.

Cars that can drive and deliver themselves left the pages of science fiction and entered reality in the 2000s thanks to many efforts, including the one at Google. (Disclaimer: I am a consultant to, but not a spokesman for that team.) By 2030 such vehicles are likely to be common, and in fact it’s quite probable they will be able to travel safely on highways at faster speeds than we trust humans to drive. They could also platoon to become more efficient.

Their ability to deliver themselves is both boon and bane to rail transit. They offer an excellent “last/first mile” solution to take people from their driveways to the train stations — for it is door to door travel time that people care about, not airport-to-airport or downtown-to-downtown. The HSR focus on a competitive downtown-to-downtime time ignores the fact that only a tiny fraction of passengers will want that precise trip….

The cars won’t beat the train on the long haul downtown SF to downtown LA. But they might well be superior or competitive (if they can go 100mph on I-5 or I-99) or the far more common suburb-to-suburb door to door trips. But this will be a private vehicle without a schedule to worry about, a nice desk and screen and all the usual advantages of a private vehicle.

There are many practical reasons why this isn’t going to be a workable solution in 2030. First off, thanks to peak oil a reckoning is looming. Either gas prices become unaffordable, meaning few people will be able to afford a self-driving car, or people start seeking out alternatives to lighting fossil fuels on fire to get around. True, those alternatives will to an extent include an electric car. But the top range for mass market electric vehicles right now is under 100 miles. There’s a really long way to go to get them ready to drive 400 miles on a single charge. And of course, you’d need some way to power the electrical load needed to charge all those vehicles.

Second, assuming you either ignored or solved peak oil, where do all those self-driving cars go? California’s freeways are already jammed. Expanding them to handle the load of population growth would cost about as much as HSR – $25 billion to expand Highway 99 in the Central Valley alone. Expanding freeways in the urbanized areas is not only very costly, it’s a bad use of precious urban space.

Third, by 2030 many Californians will be living in urban centers and trying to drive much less than they do now. The evidence shows not only that Millennials are driving less, and Census data suggests Boomers will drive less too as they age. Sightline Institute has this to say about the year 2030:

we expect per capita driving for the adult population of the United States to fall by 6.5 percent through 2030.

Yet these geniuses at Singularity University think that self-driving cars will be all the rage in 2030. Uh-huh.

Templeton gives himself away when he writes of “the usual advantages of a private car” – which are what, exactly? Especially on a trip from SF to LA, being in an automobile just sucks. It’s small, cramped, you can’t move. Same on a plane, really. On a train, however, the trip is far more comfortable. Tired of being in your seat? Rather than being stuck or having to stop the car, you can stretch your legs by walking to the cafe car, getting a drink, sitting down to watch the view, then walking around some more before returning to your seat.

A train is even better if you have kids – rather than having them being stuck and antsy in a car, they have more room to enjoy themselves on a typical train.

Templeton’s desire to breathe life into 20th century methods of travel extends to the airplane, which like the automobile is older than the bullet train. Here again, facts get in the way of his unrealistic dreams:

The air travel industry is not going to sit still. The airlines aren’t going to just let their huge business on the California air corridor disappear to the trains the way the HSR authority hopes.

Actually, that’s precisely what they are going to do. Ask an airline exec yourself. JetBlue has embraced HSR and wants out of the short-haul markets where most airlines don’t in fact make much money. No other airline lifted a finger to block California HSR. Airports across California are strongly supportive of HSR, indicating they understand HSR isn’t a threat to their business but will help open up slots for the more lucrative medium- and long-haul flights.

Despite these facts, Templeton plows right on:

These are private companies, and they will cut prices, and innovate to compete. They will find better solutions to the security nightmare that has taken away their edge, and they’ll produce innovative products we have yet to see. The reality is that good security is possible without requiring people arrive at airports an hour before departure, if we are driven to make it happen. And the trains may not remain immune from the same security needs forever.

They can’t cut prices if the underlying realities of rising oil prices won’t let them, and an electric plane is quite far from being possible. The FAA has projected high airfares for the rest of the decade at least. As to security theater, I’ve not seen the airlines make a move to deal with this growing threat to their industry yet. I wouldn’t be surprised if their solution is to saddle the competition with it anyway.

Templeton really wants to preserve the 20th century, going to great lengths to propose ways to keep air travel as a primary method of travel between SF and LA:

The fast trains and self-driving cars will help the airports. Instead of HSR from downtown SF to downtown LA, why not take that same HSR just to the airport, and clear security while on the train to be dropped off close to the gate. Or imagine a self-driving car that picks you up on the tarmac as you walk off the plane and whisks you directly to your destination. Driven by competition, the airlines will find a way to take advantage of their huge speed advantage in the core part of the journey.

What Templeton ignores is that a downtown HSR station is actually more convenient for most people in California than most urban airports. More people live and work near Transbay Terminal than SFO. SoCal’s airports are generally on the fringes of the urban area and not very convenient for those living in the densest populated places, especially in LA. A self-driving car on a tarmac? Really? Indulgent, and perhaps viable, but it’s not a solution for the masses. Nor is it necessary when we’ve seen HSR work so well around the world.

Even the mythical electric airplane also suffers from the same problem as the electric car: where do you put it? The cost of expanding airports to handle more short-haul flights is well into the tens of billions. None of these technological transportation solutions are cheaper than HSR.

And of course, none of them carry the smaller carbon footprint of HSR. Anyone projecting what the world of 2030 will need who doesn’t once mention climate change or carbon emissions is not to be taken too seriously as a futurist.

Templeton makes a nod toward maglev, which is fine, at least that’s a train which fits the world we can expect in 2030: an urban society, with little use for driving, where oil is either way too expensive or where better alternatives have been provided. But he slides back into his 20th century defending ways when he touts teleconferencing as a magical transportation solution:

Decades after its early false start, video conferencing is going HD and starting to take off. High end video meeting systems are already causing people to skip business trips, and this trend will increase. At high-tech companies like Google and Cisco, people routinely use video conferencing to avoid walking to buildings 10 minutes away.

Telepresence robots, which let a remote person wander around a building, go up to people and act more like they are really there are taking off and make more and more people decide even a 3 hour one-way train trip or plane trip is too much. This isn’t a certainty, but it would also be wrong to bet that many trips that take place today just won’t happen in the future.

It’s true that video conferencing will eliminate the need for some business trips, but by no means all. And even as urban Americans are supposedly interacting solely through their smartphones and social networks, ridership on mass transit in cities continues to climb. In 2030, people are still going to want to travel, and they’ll need a convenient, comfortable, non-oil based way to do it. Nothing in Templeton’s article has suggested there’s a better answer than HSR.

Besides, telepresence robots? Seriously?

Once you get to the end of the article, though, you learn what this is really all about. Templeton just won’t accept that the 20th century model of automobile-based sprawl really is going away. He just refuses to accept the possibility:

Like it or not, sprawl is increasing. You can’t legislate it away. While there are arguments on both sides as to how urban densities will change, it is again foolish to bet that sprawl won’t increase in many areas. More sprawl means even less value in downtown-to-downtown rail service, or even in big airports. Urban planners are now realizing that the “polycentric” city is the probable future in California and many other areas.

This is just blind assertion without facts. Most of the SF Bay Area has urban growth boundaries. As SB 375 continues to be applied in practice, it’s going to get a lot more difficult for counties and regions to promote sprawl. But it ultimately won’t be up to the laws of the legislature. Sprawl is the product of three factors: favorable laws, cheap oil, and cheap credit. Even with credit being quite cheap these days, the end of cheap oil has cut sprawl off at the knees. Growth in the exurbs has fallen off dramatically. Companies are increasingly moving their headquarters from office parks to office towers. Twitter is in downtown San Francisco just a few blocks from the Transbay Terminal, whereas Facebook appears determined to revive the office park and is taking a lot of criticism for it.

Companies like Twitter and Amazon (which is poised to build three new office towers in downtown Seattle) understand that their employees don’t want sprawl, they want the city. That’s another trend Templeton has discounted in his drive to maintain the failing 20th century model.

High speed rail is a model whose success has repeatedly been proven around the world. It’s modern and high tech and while that may not be the kind of dramatic brand-new tech that some might want, at least it actually meets what will likely be the needs of California in 2030, rather than trying desperately to maintain 1980 forever.

  1. joe
    Sep 4th, 2012 at 21:34

    I saw a Google Self driving car on Castro St. nearly get hit by the Caltrain gates as they were closing for a on-coming train. I believe there was an intervention to get it to surge under the gate.

    Here is a repeat of the Charlston Road accident.
    A woman was killed on the Caltrain tracks Friday afternoon when the car she was driving was struck by a train shortly after 5 p.m. at the Charleston Road crossing.

    Simon said the car was crossing the tracks during peak traffic. It appeared to get stuck behind the backed-up traffic and remained on the tracks when Train #369 struck. He said the northbound Express train was on its way from Mountain View and was rolling at full speed — 79 mph — when the collision occurred.

    Richard Mlynarik Reply:

    I saw a Google Self driving car on Castro St. nearly get hit by the Caltrain gates

    Google “Project Turing” mission objective: +1.

    joe Reply:

    Yes, tuning. It is fixed by tuning. I think the driver had a knob and he tuned the vehicle. Or was it a tuning fork? Just a patch to the software. Maybe add a sensor.

    joe Reply:

    I was just thinking, CEQA should be left alone. No waivers.

    Safety critical software, like this self-driving car. I bet Richard would be okay with Google by passing the standards used for approving safety critical software.

    Safety-critical software means that failure of the software will very probably result in loss of human life. Providers of safety-critical systems exercise extreme caution against system failure. Additionally, most industries are closely regulated by government agencies with stringent certification requirements

    Paul Druce Reply:

    That wasn’t a typo. Turing test is whether a computer is indistinguishable from a human to a blind observer. This one has successfully mimicked the stupidity of normal human drivers.

    joe Reply:

    Yes, I missed that completely. In fact it was funny. The car was able to screw up like the motorist.

    John Nachtigall Reply:

    So a self driving car did not get hit and one driven by a human did. You realize that your examples argue for not against a self driving car

    adirondacker12800 Reply:

    I’m sure they have programmed the self driving cars to pay attention to the warning signs that a railroad crossing is coming up and not enter the railroad ROW until there’s enough room on the other side to exit the railroad ROW.

    joe Reply:

    From what I saw, I couldn’t tell.

    My guess is the car didn’t have a sensor – it did not behave as if it sensed the gates were dropping. It slowly proceeded onto the tracks as the gates were dropping and nearly had it’s head knocked off.

    I know that google makes a lot of software but how little if any of that that is critical software. IMHO it’s far cry from a demonstration to fielding software for vehicles to having the software operate a vehicle, not advise a driver.

    Nathanael Reply:

    This is exactly the sort of thing which the Google programmers will fail at; they’ve never written a safety-critical system in their lives, most of them.

    As I have said before, robot cars are probably safer than the average human driver, but they won’t be accepted unless they are *many times* safer than the average human driver.

    And that, they will not be. One of them will do something stupid like driving under a crossing gate, and they’ll be illegal again.

    joe Reply:

    The Charleston Crossing death was an out of town visitor confused by a bad grade crossing. It illustrates the kind of problem the Google car autonomously created for itself at far safer crossing.

  2. D. P. Lubic
    Sep 4th, 2012 at 22:18

    There are a couple of things that I note are missing in Mr. Templeton’s argument:

    (1). You will still need a road system with self-driving cars. If the cars are electric or some other alternative power source, how do you pay for the road system without the existing fuel tax? This is a question that should be addressed anytime anybody talks about alternative power sources, but it never comes up.

    (2). Robert brought this up, and I’ll repeat it–where do you put that tin can when you’re not using it? Sure, the self-driving car can go to another parking spot at some distance from where you are, but there still needs to be that facility, and there needs to be a way to pay for it. Incredibly, some people have argued that self-driving cars may not have to park, but can just cruise around waiting until either a parking spot opens up or even can cruise until you’re ready for the car to come back. Can you imagine what that scenario would do for street congestion? How come these smart guys forget such things?

    (3). Finally, cars do have limits on longer trips. I should know, I’ve got a decent car (Toyota Camry), but even it gets cramped and tiring during a ride of two hours or more. Why do we want that?

    Why do these people, who are smarter than the average “you’re a dirty Communist” types, want to discount the comfort you can have on a train compared with automobiles and airplanes? Why do these smart people ignore the real cost of roads, as has been discussed so many times here?

    John Nachtigall Reply:

    1. Good news, we have a road system.
    2. In my mind I think of it more like the amusement park rides where there is a boarding station the cars sit at and wait to be dispatched. Since they are self driving oyu can relaly pack them in since no one has to get in or out when they are in “wait” mode. Basically like taxi stations at airports. Since you will not need parking at individual stores total amount of parking would go down. 1 walmart parking lot could hold thousands and serve many square miles of area from the lot.
    3. longer trips would not be possible unless you go gas (which defeats half the point). I think you end up with a hybrid of cars for the short trips and rail/air for the long trips.

    The real problem is keeping the cars communal considering there is going to be 2-5% of the population that is incapable of sharing (stealing, vandelism, etc.)

    blankslate Reply:

    1. Good news, we have a road system.

    Fantastic. I guess we’ll set the ongoing maintenance budget to zero because concrete is a magically eternal material.

    John Nachtigall Reply:

    as opposed to a 100% rail future…or do you think that rail systems have 0 maintenance? Any system is going to have maintenance costs.

    Brian Reply:

    Our current CA state road system has an $200 billion maintenance deficit over the next twenty years.

    HSR will maintain itself once operational and return cash.
    – It will do even better if we raise the gas tax high enough to maintain existing roads.
    – It will also do well if we don’t raise gas taxes and let I-5 and SR-99 collapse into slow, potholed, messes instead

    How do you proposed to fund that $200 billion road gap John?

    John Nachtigall Reply:

    I am going to go with the current system of a combination of gas taxes and general fund money supplemented by federal highway funds…what do I win?

    Brian Reply:

    So you vote for significantly raising the gas tax. Good to know.

    That will make HSR and transit perform much better than today’s projections.

    Nathanael Reply:

    Specifically, do you propose raising the gas tax (which will make rail far more popular), or do you propose cross-subsidizing highways out of the general fund for $10 billion a year?

    The latter is not really a sensible use of money, is it? $10 billion a year would get you a huge amount of rail service.

    John Nachtigall Reply:

    I forgot to mention…it is far from assured that HSR will pay for itself…that is the plan, but it is not proven yet. However, I am sure that local rail (BART, etc.) does not pay for itself and that is what is analogous to local roads.

    So how are you going to pay for both the capital investment in a whole new local rail system that goes everywhere (to replace these evil roads) as well as its maintenance.

    synonymouse Reply:

    @ Brian

    The 3 crones and the drone will find your $200bil, with GOP collusion, because they know roads are popular with the electorate, you remember, the people who vote for them and keep them in perks and rides on Air Force One. Boxer always finds funds for new lanes on 101 in Marin.

    The crones are like Joan Rivers, preserved.

    blankslate Reply:

    Nice non sequitur. The question was how roads will continue to be maintained when the only dedicated source of funding (gas taxes) declines due to alternative fuels, your answer was that we “already have a road system.” I don’t know what rail system maintenance has anything to do with this discussion, but since you decided to dodge the question and bring it up, here are my answers: 1) rail systems are cheaper to maintain, 2) the source of funding for their maintenance is not the gas tax.

  3. John Nachtigall
    Sep 4th, 2012 at 22:28

    I think you aree taking this all wrong. The vast majority of trips are under 20 miles. HSR does not compete in that area. I logical extension of the self driving car is that it becomes a communal resource. Instead of everyone owning their own car they are more like a resource that picks you up and drops you off when you need it. You will still need quite a few for the rush hours, but net much less then we have now.

    Using myself as an example, on the average workday I use my car for less than 20-30 minutes a day. So for 5% use I own and park and use all the resources to have my own car. If I could be guaranteed a ride to and from work at my own chosen time I would have to reconsider the need for a car.

    And they don’t have to be gasoline based, since it could be a central fleet use of electrical vehicles would be much more practicle. I could never own an electrical car because I have no place to plug it in since I rent.

    I did not see the article as anti-rail. I think self driving cars could really compliment your rail future, not endanger it. You have to face the reality that people are not going to be able to walk or bike 20 miles every day which represents their basic transportation needs

    D. P. Lubic Reply:

    I concur, John, but the author of the article plainly wonders if conventional (if high speed) railroads are going to be able to compete with self-driving cars. It’s right at the bottom of the first page.

    I find it particularly interesting that he predicts these cars could go 100 mph. I can personally tell you that energy/fuel consumption goes up quite dramatically past 70 mph or so. That’s why I question the idea of regular driving at those speeds, even though some do that in Germany.

    Part of the reason for this is that a car is comparatively short and wide. Even the best aerodynamics can’t completely compensate for the difference in effect when compared with a body that is very long and skinny, which is what a train is.

    Miles Bader Reply:

    The “OMG self-driving-cars will solve all problems!1!” meme is just the usual “tech-fix” delusion at work. Many people have a nostalgic and ideological attachment to cars, and desperately want some solution to the myriad problems they cause which is “still a car,” so they’ll desperately grasp at any straw that pops up. Despite the essential failure of the 1950s suburban dream, there’s an attitude that “just this one change will make everything work great!”

    Self-driving cars are neat, but they aren’t a replacement for better designed cities, urban mass-transit, bicycling, or walking—much less HSR.

    Andy M Reply:

    The question is, is a self-driving car still a car if it becomes communal and hence maybe you can’t take it outside its home area and have to book or call it prior to use. It sounds more like a taxi to me. I can see strong advantages to the communal approach. for example you wouldn’t need to bother about parking but could just get off and the car would return empty to the charging station or go and meet its next customer. In fact it would be not going communal that would be insanity. But once you’ve got people accepting the communal aspect, you get a convergence going on with the differences between cars, taxis and public transit becoming increasingly permeable.

    Neil Shea Reply:

    Think Johnny Cab from Total Recall

    John Nachtigall Reply:

    If I call them self driving mini-trollies would you object?

    A communal pool of self driving cars would act just like a trolly system without the restriciton of having to lay down tracks. As Andy wrote above this it acts like a taxi/public transport.

    I would think a future of self driving cars for the sub 20 mile trips and HSR for long distance you appeal to the supporters of HSR on this board.

    THe author admits it may be a boon or a bane, no one really knows how people would react to the advancement. We will probably all live to see how it plays out which is exciting.

    D. P. Lubic Reply:

    “If I call them self driving mini-trollies would you object?”

    Actually, something like that has been around for a long time now, although it turned out to be frighteningly expensive with the technology of the time. It’s what’s called GRT, which stands for Group Rapid Transit. This is an on-demand service with relatively small automated vehicles. The best known example may be the Morgantown, W.Va. “PRT” system, dating back to the 1970s.

    I’ve ridden that system, and technologically it’s impressive, and has an excellent safety record, but I think the designers should have gotten over the idea that steel wheels on steel rails was outmoded. The switching is done by changing what I call a gripper on one side or the other of the cars, which I think needlessly complicates the rolling stock. Worst part was ride quality; the thing bucked like a wild horse. Whooee! Not the smooth ride of even 1880s vintage rail cars!,mod%3D0&um=1&ie=UTF-8&hl=en&tbm=isch&source=og&sa=N&tab=wi&ei=sIVHUJONLqi20AHu6oCgAw&biw=1024&bih=677&sei=LoZHUI3TGYrv0gH4oYD4CA

    Miles Bader Reply:

    It’s not a “trolley,” it’s a car. Like all cars, it’s extremely space-inefficient, and energy-inefficient, and these things make it unsuitable as the core of an urban transport system. Trying to play word games doesn’t change that.

    The main problem with the U.S. is the stupid notion that a single mode suffices for all transport needs. A good transport is a mixture of modes, walking (yes!), bicycling, rail, buses, car… and just as importantly, “non-transport” solutions: arranging the structure of the city so that people have to travel less and can use less expensive modes.

    Self-driving cars may indeed prove a good solution for a certain part of the transport space, essentially making a “taxi” less expensive and more convenient. But they aren’t a magic bullet that will save the U.S.’s dysfunctional urban design and transport planning from itself. Unfortunately, all too many people (especially nutters with ideological attachments to private transportation) seem to be treating them as such …

    John Nachtigall Reply:

    here is a thought that is going to blow your mind.

    Instead of bending the transport system to your vision of urban design…we bend urban design to the technology of the transport system. Oh wait, we already have. Suburbs, exurbs, etc.

    adirondacker12800 Reply:

    If the zoning only allows suburban development take a guess what people will build.

    John Nachtigall Reply:

    it is not the zoning laws that prevent urban development, it is cost/benefit. You can get a much larger house for a much smaller price in the suburbs. it is not zoning.

    adirondacker12800 Reply:

    If all the town allows is single family houses the only thing you’ll be able to buy is a single family house. When faced with the choice of living under a bridge in a cardboard box and buying a single family house most people will chose the single family house. Even though they were aiming for a three bedroom in a garden apartment building. Or a 1500 square foot loft.

    John Nachtigall Reply:

    There is plenty of urban housing. Most people in the US just don’t want to live in a shoebox. There was just a huge condo boom because it was hip and cool. You can right now find plenty of opportunities to live in urban areas, I am sorry it is inconvinient to your world view, but given the choice, most people want more house for less money and choose to commute from the burbs

    adirondacker12800 Reply:

    If urban housing is so plentiful why is it so expensive?

    Jon Reply:

    John Nachtigall – you haven’t tried to rent in San Francisco recently, have you?

    A quick trip to tells me that the average rent for a studio in SF is $1250. That sounds about right. If no one wants to rent shoeboxes in urban areas, why are they so expensive?

    John Nachtigall Reply:

    $1250 is not expensive

    I pay $1850 for a 2 bedroom in Santa Rosa (an exburb) 1 hour away

    besides, the very thing you like about urban living (high density) keeps the prices high. You can only fit so many places into so many square feet. This is not Hong Kong where you can sell a 50sqft space as an appartment. That is not zoning, that is basic safety and livability standards.

    Jon Reply:

    Well, let’s compare like with like. The average rent for a 2 bedroom in my SF neighborhood is apparently $2800. The average rent for a studio in Santa Rosa is apparently $775. In short, SF is more desirable to live in than Santa Rosa. Who woulda thunk it?

    In theory higher density living should be cheaper, because increasing the number of apartments on any given acre of land means more income per acre for the landowner, so rents per unit can be cheaper. But in reality, the desirability of living in urban areas plus the shortage of urban housing caused by pro-suburban zoning laws means that rents are higher than in the suburbs.

    Case in point. As one commenter says, “Christ, SF is in the middle of the biggest housing crunch since the earthquake, and people are protesting the replacement of a KFC with a mixed-use building that encourages cycling? We are so fucked. The DR/ER review process in CA and SF needs to be seriously reformed.”

    Alon Levy Reply:

    John, so which is it – “not expensive,” or “keeps the prices high”?

    joe Reply:


    Rents and desirability are not equal.

    I prefer Montana over SF but the salary is not that high in MT so the rents are less and people accept less to live there and enjoy the benefits.

    Also SF has unique attributes beyond density. Neighborhoods like Noe Valley are not legal to build anymore – you cannot builds homes that close with old growth wood and without adding parking.

    Nathanael Reply:

    No, John, it’s the zoning. You haven’t done your research on this.

    Jon Reply:

    Rents and desirability are not equal.

    Yes, in a free market they are. The rent represents the price someone is willing to pay. How is that based on anything other than desirability?

    Neighborhoods like Noe Valley are not legal to build anymore – you cannot builds homes that close with old growth wood and without adding parking.

    Yes, they’re not legal to build because of zoning codes restricting high density development. Requiring that all new properties have parking is one of the main ways in which zoning laws prevent density. You’re making my point for me.

    John Burrows Reply:

    I know that the mere mention of San Jose can unduly excite some of us—But here goes.

    On Sunday I noticed that a 265 unit condo development next door to where I live has sold out. The sales office is now for lease. Across from the nearest Safeway a whole block has been fenced off for a 5 story mixed use development. Two blocks in the other direction construction started up again last week on a 250 unit condo project that was abandoned in 2008, although this time it will be rental.

    And downtown, for the first time in several years, two new high-rises, totaling nearly 800 rental units are in the pipeline. In fact the comment was made that the completion of these high-rises might be what it takes for downtown San Jose to reach critical mass—Maybe a little optimistic.

    So from what I can see urban living is alive and well, and in San Jose of all places.

    John Burrows Reply:

    And for urban planners who think that you can’t successfully build residential next to the tracks—Some of the units in the sold-out condo development back up to within less than 30 feet of the VTA tracks and less than 70 feet of the nearest Caltrain track.

    blankslate Reply:

    Indeed, the notion that suburban living is “what everyone wants” is really quite laughable to anyone who takes even a cursory look at what has been going on in the South Bay recently. Minimum 4-story multifamily housing near transit is practically the only type of residential that has come through the development pipeline in the past five years.

    Neil Shea Reply:

    Hey John, be careful you’ll rouse the San Jose haters for whom everything is fixed in the past. You’ll spot them when they spell San Jose with an accented e. Remember, good paying jobs, dense new housing, low crime rate and many transportation options do not make a good city; being San Francisco does. SF’s aura of greatness extends to portions of the E Bay but never to the S Bay. Just so you’re clear…

    Joey Reply:

    The city is changing for sure but it will take a long time to change the land use patterns of the last 50 years, on which the entire valley is based. As it is, you’ve got both massive residential and job sprawl (and I don’t see any particular momentum toward fixing the single-story office park problem). Diridon station underperforms compared to other CalTrain express stops (whereas Transbay would double the number of jobs accessible by CalTrain).

    And I’m not sure what good transportation options you’re talking about, unless you mean expressways. VTA is a joke as it is, and there seems to be no political will to improve it. BART will allow access to the east bay but not solve the problem of car-centric intra-valley transportation.

    Conclusions: Is SJ worth paying attention to? Yes. Does it deserve HSR service? Yes. Will it densify going forward? Somewhat. Will downtown SJ ever surpass downtown SF? Probably not in our lifetimes.

    blankslate Reply:

    The city is changing for sure but it will take a long time to change the land use patterns of the last 50 years, on which the entire valley is based. As it is, you’ve got both massive residential and job sprawl (and I don’t see any particular momentum toward fixing the single-story office park problem).

    There is some momentum. Most new office development proposals of the past few years have been 4-6 stories or higher, concentrated in North San Jose, Santa Clara, Sunnyvale, and Mountain View. I’m not aware of any single story proposals in the past few years.

    Granted, these are still office parks and they still have wayyyy too much parking, but it does show that the demand for density is rising, even in the employment sector.

    Diridon station underperforms compared to other CalTrain express stops (whereas Transbay would double the number of jobs accessible by CalTrain).

    Downtown SJ is not where it’s at in the South Bay. Probably never will be. The Caltrain stops between Lawrence and Palo Alto, as well as the ACE/Capitol Corridor stop at Santa Clara-Great America, are the most popular rail stations. From my eyeball estimate (daily CC rider), about three times as many passengers board ACE and Capitol Corridor at Great America as Diridon in the afternoon.

    Will downtown SJ ever surpass downtown SF? Probably not in our lifetimes.

    Probably not ever, but that’s ok. SJ & other South Bay cities should work towards densifying, activating key streets like El Camino Real and North First Street with mixed use development, and speeding up their crappy light rail & bus lines. I believe that there IS political will to do all these things, and market forces that will keep increasing the political will (e.g. more people will move to areas near transit because that’s the only place housing is getting built, they will eventually demand better transit). It will get better. Let SF be SF and SJ be SJ, and everyone will be happy.

    Richard Mlynarik Reply:

    There would be few things around here better than massive density increases in San Jose — massive improvements in Oakland being one of them.

    That’s not the problem.

    The problem is tens of billions of dollars of public finances — regional, state-wide, federal — being quite literally flushed down the crapper to support a density that doesn’t exist now, won’t exist tomorrow, and might only be showing the beginnings of a hint of existing 50 years from now.

    So SJ should knock itself out with the 4 story condos and get completely excited about replacing one empty parking lot with a Chick Fil A franchise. Great stuff. Moving in the right direction. Keep it up! But it’s not stuff that today even remotely justifies the massive and criminal distortion of public expenditure across the entire state. (And the country for that matter: the $20bn or so of pure unadulterated waste that the combination of BART to the SJ Flea Market and HSR to Los Banos represent is a significant piece of federal transit and HS even on a multi-decadal scale.)

    joe Reply:

    Dude who flushes his toilet with water from the Hetch Hetchy Reservoir continues to find fault with infrastructure investments in not-in-my-back-yard CA.

    Joey Reply:

    What gave you the impression that this had anything to do with NIMBYism?

    Nathanael Reply:

    Step one would be to make density legal, as the people who’ve studied the zoning codes have pointed out. The rules in Seattle are absolutely absurd; there’s an entire essay on how they result in “zoning-based housing design”.

    Winston Reply:

    There is likely a market for self-driving taxis. However, if they start eating into the market share for regular cars, they will start suffering from many of the same problems that regular cars do; for example, the demand for them is likely to be very strongly peaked during commute hours, meaning that a lot of them will have to park for the bulk of the day. Further, since they would need to be available on demand they would have to park fairly close to where they will be needed, so you haven’t really solved the parking problem that cars create in cities. Also, a big issue with something like an automated taxi is going to be cleanliness. I wouldn’t be looking forward to following after some guy who just took a crap on the seat.

    Andy M Reply:

    The pleasant thing about travel patterns in cities is that they are highly preictable. Not on a person by person basis but as an aggregrate. You know at this time you will need this many cars on the road, give or take a handful. So you don’t need to park those cars in expensive downtown locations but can have them self drive into the cities from their peripheral maintenance/charging stations to have them arrive in the right locations precisely when the aggregate demand rises.

    Of course, if you think about it, that is precisely what a bus or commuter train does. These rarely park up all day in prime downtown locations but are scheduled to be there when needed. Thus we have another step towards that convergence I mentioned earlier. Only the bus or train is much more fuel efficient, is much more comfortable and spacious, etc, and so once you’ve sold the micro solution to people accustomed to their own cars, it’s but a small step to upsell them to proper public transit. Call them super-HOVs if you like, with an in-built real-time social networking function.

    What do the cars do between peak times? The same as buses or trains do. Some remain in service, with maybe some sort of yield management fare system encouraging off-peak travel. Others undergo maintenance or similar. Remember electric cars have longer charging cycles than the cars that burn the remains of dead plants and animals that somebody took out of the ground. I don’t drive a car myself, but often my work colleagues need to take some time off work to bring their car to the garage to have something fixed. I never need to bring my bus to the garage because the driver does that for me.

    Winston Reply:

    You still have the fundamental problem that you have to park a big portion of your vehicle fleet for a substantial amount of time and that those vehicles take up a lot of space compared to transit vehiches. If you store your fleet downtown, you haven’t accomplished anything wrt parking. If you park them a distance away from downtown then you’ve solved the parking problem, but you’ve made traffic dramatically worse and in an urban environment with a variety of road users you can’t do any capacity increasing tricks like platooning to get more out of your road space. Yes, there is a future for automated cars and it’s a bright one, but they have lots of very significant problems.

    Andy M Reply:

    Have you ever been to a bus garage or a train shed? The big difference to a normal parking lot is that buses are parked bumper to bumper. You don’t need to allow access paths to every individual vehicle but can have a last-in-first-out system, or even a first-in-first.out. A parking lot for private cars, however, is random access. How much space is required to assure those paths that are required to enable the random access functionailty. Maybe you’re approaching half the total space? If you’re efficient, maybe a third. But it’s still significant. So if you replace private car parking on a lot by lot basis as the market swings, you are actually releasing downtown space for other uses.

    Of course public transit is way more efficient still. But I’m thinking of this as more of a first step in guiding people towards transit.

    Wdobner Reply:

    Why restrict yourself to self-driving cars when there’s NASA LaRC’s <a href=""Puffin proposal? Autonomous airplanes are likely to occur before autonomous automobiles. Why restrict yourself to the road when you could hop in a self-flying aircraft?

    The Puffin is particularly well suited to the CHSRA system as its range is merely 50 miles. If a home-base for a number of the autonomous aircraft were co-located with an HSR station it could allow access anywhere within a 1900 square mile catchment area in minutes without recharging, and an area of 7800 square miles with a recharge to return to base. The HSR system provides the backbone for the long-distance travel, while the autonomous aircraft allow fast access to sparsely populated areas along the route without a significant travel time penalty.

    Wdobner Reply:

    Crap, screwed up the HTML. It’s the Puffin from NASA Langley.

  4. missiondweller
    Sep 4th, 2012 at 22:57

    The author admits that HSR does better than cars over long distances.

    Well, that’s what its proposed to be used for!

    If there were a HSR line between Stockton and say Oakland, Stockton would not have gone bankrupt. Why? Because Stockton has a glut of housing while the Bay Area has a shortage. If you could buy a cheap home in Stockton and easily commute to your job in SF you would have the benefits of both a high paying job and the affordable home. Stockton would have maintained their property tax base.

    All that, and no need for urban sprawl. Its a hypothetical example but hopefully makes a point.

    Alon Levy Reply:

    It’s worth asking why the Bay Area has a housing shortage.

    Matthew B Reply:

    It’s not for lack of land, given that LA has higher metro area densities than the Bay Area. Tokyo and Paris manage to maintain very high quality of life despite even higher residential densities.

    VBobier Reply:

    It’s cause high density is seen as a place where poor people live and poor people bring crime, at least from what I’ve read. this isn’t always true of course, I live in a poor area and crime isn’t any higher here than an area that has higher incomes, so I don’t see higher density as a real problem, just something some don’t want near them… But then bigotry is alive and well…

    Nathanael Reply:

    Because, basically, it’s not legal to build residential up higher.

    synonymouse Reply:

    @ Missiondweller

    Interesting you chose two cities historically notorious for violent crime. Curious likewise that in Southern California route planning was politically manipulated to deviate to a poorer area, Palmdale and the Antelope Valley, whereas in Northern California it was gerrymandered away from the center and east Bay, Oakland and Stockton, to a richer area, San Jose and Silicon Valley, PAMPA. Strange, I don’t have a facile explanation for that.

    Winston Reply:

    In L.A’s case they have really, really wanted HSR to Palmdale for a long time since that would open up a big area for development and potentially solve L.A.’s airport problem. In the case of the Bay Area, San Jose has a serious case of tower envy and wants big city amenities. As for running trains up the peninsula, that was guaranteed as soon as the Bay Area decided to build their signature bridge to Oakland.

    We would all have been better off if instead of spending $6b to build a new eastern span of the bay bridge, we had spent $2b retrofitting it (the costs of doing so were estimated at $1b) and returning it to its 1940’s configuration (2 lanes + 4 tracks on the lower deck) and 6 lanes (or 5 now that we like bigger lanes) on the upper deck. This would have left $4 billion to build, say, a Geary street subway and a matching east bay subway and leave a perfectly wonderful bay crossing available for HSR and commuter trains. HSR trains could then continue to San Jose via upgraded Caltrain tracks and could reach the CV by traveling across the bridge and then on a new ROW roughly paralleling BNSF’s current ROW to Stockton via Antioch. Such a ROW, if done right would have also created huge time savings and a better southern endpoint for the Capitol Corridor, allowing it to be significantly more useful. It would also be much cheaper and easier to build than any of the mountain pass routes that we like to debate here.

    synonymouse Reply:

    Very interesting proposal and, or course, anathema to the SP. That’s why you have BART broad gauge and 3rd rail and, to a lesser extent, the Transbay Tube, which would have been more difficult to accommodate FRA-AAR heavy rail than the Bay Bridge.

    I agree with you things could have been done quite differently and ultimately much better. Standard gauge being the most important change, as it is possible to do both overhead collection and third rail on the same system.

    The problem with the bridge, apart from the fact the Division of Highways wanted the lanes for autos, was the harmonics limiting the top speed to, I believe, 35mph. The Tube is vastly superior in that regard, plus I suspect the maintenance is much less. The Tube is BART’s major engineering feat.

    Had the subway bond issues passed in 1937 obviously BART would have had a much more difficult problem with Market Street and Geary would have been likely secured against Urban Removal in 1956. BART would have probably wanted Mission St., but a subway-surface operation there would possibly have shunted BART to the route you envision, the current Caltrain Peninsula ROW. Especially if the Peninsula had opted to join BARTD.

    The loss of the Mission St. streetcar lines and the #40 to San Mateo was most unfortunate and I think the worst mistake of the great Muni busstitution of 1946-50. How about this – seeing as how Rose Pak’s godawful Central Stubway proves it is now permitted to blow a fortune on personal pet projects I wanna suggest my own. Let’s relay rail on Mission and/or Valencia and restore the long-lost #11 to Dolores Park and Noe Valley. Richard, what do you think?

    synonymouse Reply:

    And as to the mountain pass debate I believe the difficulties of the Golf Course Route at Tejon have been grossly exaggerated and that the price is appropriate for what you get – the optimal alignment. That was Van Ark’s epiphany – he was just light years ahead of Moonbeam.

    Winston Reply:

    The transbay tube is a very nice piece of infrastructure and I’m glad it got built. What I’m saying is that even as late as 2002 we could have done what I’m proposing. At this point it was clear that the bay bridge had way more capacity than San Francisco’s freeway network does and that returning trains to the bridge could have been done with a small impact on traffic.
    Retrofitting the bridge, returning part of it to rail and restoring part of the key system plus building the much needed Geary subway would have been a much better thing than what was actually done. As for what kind of subway to build, I’m agnostic on that point – the best thing to do would depend on the details of what you were planning to do in the East Bay.

    Joey Reply:

    If rail was returned to the bridge, the configuration probably would have been a new deck below the current deck on the west span. Maybe not particularly cheap, but cheaper than the alternatives, including the second tube which will eventually be needed.

    William Reply:

    On the Bay Bridge East span part, the retrofit, according to Caltrans, would take away one shipping lane due to the need of additional supports, last half as long as a new bridge, and cannot withstand magnitude 8 earthquakes as expected on San Andreas and Hayward faults.

    Richard Mlynarik Reply:

    William, as you, personally, very well know. MTC’s executive staff (especially its Executive Director, who is personally responsible for the Bay Bridge East Span fiasco, to the tune of five billion dollars) are consistent, habitual, and egregious liars.

    A rail-capable bridge was technically and financially completely feasible, with no compromise whatsoever on the shipping lanes. MTC’s executive staff, abetted by ever-transit-hostile Caltrans, and in active collusion with the “competitive” team “designing” the “signature” span, decided that no such thing should be constructed or considered or be allowed to be considered.

    Anybody who claims otherwise is simply an outright liar. I was there throughout the so called “Bay Bridge Design Task Force”, and saw it all of it — at least that small part that they mistakenly allowed to be seen in public — go down.

    missiondweller Reply:


    Stockton has a housing surplus which BAY Area workers would benefit from.

    Oakland is a point where several BART line converge, allowing commuters to transfer from HSR to BART to complete their commute.

    It was just an example to prove a point. Perhaps a real planner would have a better idea. I’m a finance/economics guy.

  5. ZStern
    Sep 4th, 2012 at 23:21

    I read this article earlier today and I am glad to see you picked it apart Robert. I also think the technology argument is pointless. You can only wait so long until you have to build/buy something. If you continually wait for the latest Iphone you won’t ever have an Iphone…. see: US and HSR.

    Miles Bader Reply:

    Wait, is this Zalman?

  6. Howard
    Sep 5th, 2012 at 00:33

    First will come driverless trains, second driverless busses and last self driving taxis-cars. Trains will be first to go driverless because the technology already exist (its easy when the tracks steer the train). First BART will be converted (it was originally designed to be driverless), then Caltrain will be converted as part of its modernization, CHSR will be from the start, with in street running light rail the last trains to covert. The money saved not paying drivers would lower fares and increase service. Busses will take longer to go driverless then trains because they need to steer in traffic but they will covert faster then cars because they only need to learn how to drive a fixed route, not any road. BRT systems with in road sensors to guide the busses are being developed now. Most of the cost of bus transit is the cost of paying the bus drivers. When we save most of that money, after all the bus drivers are laid off, imagine how much bus transit could be expanded an no additional cost. After that will come driverless taxis, which would also become very cheap. Driverless transit could become more attractive than driving.

    Andy M Reply:

    Yes, and driverless buses would be able to align more precisely with the loading platform, so lessening the gap and making it easier to board. Also imagine the consequences for vehicle design if you no longer need to have a driver’s seat at the front. The front door wouldn’t have to be at the very front either if there was no driver, so you could have a row of seats along the full width of the front that would be popular with tourists.

    Nathanael Reply:

    Driverless trains are already used in several places — Docklands Light Rail and Vancouver SkyTrain are examples.

    The catch is, *people will not allow them to be used where there are pedestrian grade crossings*.

    The result is, there will never be widespread deployment of driverless buses or cars. Why? Because roads have pedestrian grade crossings. The first driverless vehicle which kills a pedestrian will be the “killer robot car”, and they will *ALL* be banned *IMMEDIATELY*.

    Andy M Reply:

    Someboedy once said the automobile would never catch on because where would you recruit all the men to walk in front with a red flag? Isn’t this a similar argument?

    D. P. Lubic Reply:

    The Washington Metro already has driverless capability, and in fact would normall operates as such but with a human aboard to monitor the train and passengers and to make announcements. There actually was an incident some years back where an operator left the train to go to the restroom or something, but he left his key in the control compartment. The train, responding to signals from the central operation computer, took off by itself, making all the stops and starts, picking up passengers just as it should have done. Nobody noticed for some time that there was no operator on board, at least not at the rear of the train.

    It was an interesting demonstration of the capabilities of the Metro’s automation system. Unfortunately, for the operator, it demonstrated that HE wasn’t needed. . .

    The system currently is in a manual operating mode as some problems have turned up and are being corrected following a collision a while back.

    Plenty of other systems are around that are automatic in operation or nearly so, and one of them is BART.

    Andy M Reply:

    There is an amusing anecdote about London’s Docklands Light Railway. the trains are automatic with train captains being there only to press the go button. The automatic system doesn’t just drive at a fixed speed profile, but compares the actual location to the timetable and so will drive a bit faster if the train is running late. The automatic driving system can also be overriden, however, and the cars driven manually. This anecdote was realted to me by a friend who worked for GEC who installed the first system. Prior to the opening and while the system was still under test a group of VIPs were given a trial run on the system. A technician drove the train manually on the first bit so the train wouldn’t stop unnecessarily at all the intermediate stations, and in doing so ended up running well ahead of its timetabled slot. The guide then decided to demonstarte the automatic system and switched this on. To everybody’s surprise, the train then drove off backwards.

    Probably to be taken with a pinch of salt. But nice story nevertheless.

    Alon Levy Reply:

    Thanks for the link to the list of driverless and partially automated trains.

    Note, by the way, that every single one of the driverless trains is disconnected from both light rail and mainline rail – subway-surface lines with grade crossings, and even fully grade-separated rail lines that have to share tracks with manned trains, always have drivers. Even the partially automated cases are segregated – e.g. the central areas of the RER A are mainline rail, but nothing apart from RER A trains runs on them, and out in the open where the RER shares tracks with the Transilien the SACEM system is not in operation.

    Andy M Reply:

    True, but this is but a first step. It’s normal they should start with the low hanging fruit. Many of these signalling system actually go back a decade or more when computers were still much more clunky than they are today. Since then, processing power and algorithms have made huge progress (as the Google self-driving car illustrates) and it should be possible to design systems that can handle a far larger number of input channels, detect obstructions on the line etc.

    Nathanael Reply:

    It’s not about ability. It’s about whether people are comfortable with it.

    The main cause of people getting hit by trains is suicide; the second-most common cause is utter, absurd stupidity on the part of the people walking in front of trains. An automated train can probably stop when it spots an obstruction quicker than a manually operated train.

    Yet, for some reason, people are more afraid of being killed by robots than by humans, to put it bluntly. I don’t expect this to change.

    Alon Levy Reply:

    Funny how there’s much less stupidity when the train station design makes it harder to walk across the track and easier to cross legally.

    Max Wyss Reply:

    Actually, as far as I read, the Paris Metro line 1 has mixed operation (manual and automatic), and Nürnberg has/had mixed operation as well. I know for sure that the Nürnberg trains do not have a formal cab, just an operating console.

    Alon Levy Reply:

    Bleh. Yeah, Paris Metro Line 1 is mixed, but that’s just a transition from manual to driverless operation and the line is going to be fully driverless soon.

    How does this work in Nürnberg? Was it also a transition, or do U2 and U3 share tracks with lines with manual operation?

    Wdobner Reply:

    It’s worth noting that in the 1979 WMATA “runaway” train the passengers’ most definitely took notice. The operator-less train ran from Brentwood Yard where the T/O stepped out to inspect a train as it left the yard, through Union Station, Judiciary Square, and Gallery Place without opening its doors. At each station the train stopped on the platform, the doors did not open, and, after waiting its programmed stop time, the train continued on to the next station. It was only when a passenger jimmied the lock to the T/O’s compartment and hit the door open button at Metro Center that the train finally stopped and the passengers were released.

    A few months after this incident WMATA activated the automated door-open feature on their trains. If this system had been in service the train would have run from Brentwood Yard down to Union Station, the doors would have opened, and the train would have sat there awaiting the operator closing the doors before it could continue on.

    D. P. Lubic Reply:

    Thanks for correcting a faulty memory–but I have to admit, the faulty memory story is better!! :-)

    D. P. Lubic Reply:

    One of the automated systems mentioned is one that is no longer operational–the London, England underground mail railway. Below is a link to an “urban exploration” site that features some amazing photos of this now closed but still largely intact system.

    Alon Levy Reply:

    In the late 19th century, there were plans in both the US and the UK to move mail with pneumatic tubes. I think there was even a vision of an intercity underground pneumatic tube system to move mail from one side of the US to the other.

    J. Wong Reply:

    BART operators do not “drive”; operation is computer-controlled. Accelleration and deccelleration are not under human control. The operator controls the opening and closing of the doors.

    Peter Reply:

    Don’t forget the Fremont Flyer…

    blankslate Reply:

    He also yells at people not to bring their bikes on the first car, even when it is practically empty, for reasons I cannot fathom.

    synonymouse Reply:

    Personal transportation will always be popular. First off it is DIY; doesn’t require any of Nancy’s union chauffeurs.

    How many TWU conductors does it take to text on the job? As many as you can get away with, sucka.

    When the new generation of Machine CaliUniParty Demos is being entirely bankrolled by corporate America and has rolled over, that’s why you might see driverless operation.

    The 3 crones and the drone will be pushing daisies first.

    Hopeless 20th century anachronism: Vegas. Look at the age of those pimping it – Adelson, Wynn, Reid.

    Nathanael Reply:

    DIY is never popular; it’s only done because it’s cheap.

    Chauffeurs are *always* popular.

    synonymouse Reply:

    If t hat were true there never would have been the phenomenon known as the “muscle car”.

  7. Jo
    Sep 5th, 2012 at 08:48

    Any crazy, fool thing to accommodate the automobile.

  8. Paul Druce
    Sep 5th, 2012 at 09:12

    Robert’s missing quite a few things here:
    1. The argument is flawed when it comes to comparing like to like; HSR targets long distance intercity while a self-driving electric car will target the rather more common trips that use the most fuel and cause the most pollution which HSR will do nothing about.
    2. HSR will do nothing to help sprawl and will likely encourage it.
    3. 2030 is two entire automotive replacement cycles away; I think it’s a certainty that they will be self-driving and given that Tesla is selling cars with 230 and 300 mile ranges, and there is technology on the near-horizon for significantly improved battery capacity and charging, they’ll likely all be electric as well.
    4. Driverless cars are entirely capable of resulting in increased VMT since they remove many of the disincentives to driving and increase the population base capable of using cars.

    Alon Levy Reply:

    Right now the California law requires a licensed driver to be in the driverless car.

    Paul Druce Reply:

    Right, but liable to change 20 years down the road if it proves itself (I’m curious as to whether they have ever been tested with night blackout conditions or manual traffic direction), and there are plenty of elderly who may have licenses but are not willing to drive themselves.

    Nathanael Reply:

    It will never be legal. I’m going to be quite blunt about that. You can’t make driverless cars good enough to keep them legal — not even in 50 years.

    *People are afraid of being killed by robots*. Somehow, people are generally OK with being killed by humans, relatively speaking — I guess because there’s someone to blame.

    The result is that *one* accident, out of, say, 10 million robot cars on the road, will almost certainly lead to the banning of robot cars.

    I don’t say I agree with it; but that is human psychology as evidenced throughout history.

    Nathanael Reply:

    The only way “automated cars” will be legal is if (1) a licensed driver is required to be in the car, and (2) the licensed driver is blamed (sent to prison, executed for murder, etc.) whenever the car’s software screws up.

    That would satisfy people’s desire for someone to blame. But then, I wouldn’t want to be in an auomated car and go to prison for the errors of Google’s programmers — would you?

  9. joe
    Sep 5th, 2012 at 11:38

    Safety critical software for cars isn’t going to be cheap. Google doesn’t write safety critical software so this is all interesting but once you start to deploy stuff that isn’t advisory, like the current safety technology, system safety becomes a problem.

    Andy M Reply:

    Software is facing safety issues in many areas. That’s why, for example, we have SIL classifications. Progress is being made all the time in coming to grips with safety. Some of this experience and knowledge will spill over into other sectors.

    joe Reply:

    Cars have lots of software, more than aircraft. Its expensive part for the automobile. Cost per line of code is far below the safety critical levels the airlines pay. When cars become autonomous and no longer advisory, the cost per sloc is going to jump or cars will be grounded pending fixes.

    For today’s premium cars, ”the cost of software and electronics can reach 35 to 40 percent of the cost of a car,” states Broy, with software development contributing about 13 to 15 percent of that cost. He says that if it costs US $10 a line for developed software—a cost he says is low—for a premium car, its software alone represents about a billion dollars’ worth of investment.

    The avionics system in the F-22 Raptor, the current U.S. Air Force frontline jet fighter, consists of about 1.7 million lines of software code. The F-35 Joint Strike Fighter, scheduled to become operational in 2010, will require about 5.7 million lines of code to operate its onboard systems. And Boeing’s new 787 Dreamliner, scheduled to be delivered to customers in 2010, requires about 6.5 million lines of software code to operate its avionics and onboard support systems.

    These are impressive amounts of software, yet if you bought a premium-class automobile recently, ” it probably contains close to 100 million lines of software code,” says Manfred Broy, a professor of informatics at Technical University, Munich, and a leading expert on software in cars. All that software executes on 70 to 100 microprocessor-based electronic control units (ECUs) networked throughout the body of your car.

    Andy M Reply:

    yes, but not all the software is safety critical. It’s a question of good design to separate out the critical and non-critical components (maybe also run them on separate processors to be totally safe). Also, cars are likely to be produced in far greater volumes than planes and so the per unit costs of development are lower.

    Matthew F. Reply:

    Unfortunately they mostly communicate via a common bus, which can open up interesting security risks….

    Andy M Reply:

    Indeed that is a challenge, but not an insurmountable one.

  10. Matthew F.
    Sep 7th, 2012 at 02:01

    By 2030, we could see a situation where self-driving cars are the vast majority of vehicles on the road. But unlike human drivers, they’re never impatient, never follow too closely, and their behavior can be optimized to maximize freeway throughput. I’m no expert but I would bet that with 100% self-drive cars, we could double freeway throughput, and vastly reduce congestion.

    But all that aside, they do not compete with HSR. They compliment it.

    Alon Levy Reply:

    Cars follow at the closest safe distance, and even closer, based on stopping distances. Since most of the stopping distance is braking distance and not reaction time, driverless cars are not going to do much for road capacity.

    Andy M Reply:

    The stopping distance is partly explained by the guy in front possibly having more powerful brakes than you. So if he stops suddenly for whatever reason you’ve got to be able to stop as well without hitting him.

    Self driving cars could communicate with one another and thus prevent such scenarios by braking at exactly the same rate. You could probably even run them fender to fender like a train which might actually even be safer as there would be less deformation on impact.

    Alon Levy Reply:

    PRT advocates have said exactly what you’re saying for a while. The technology’s not there after decades of research. Hell, some of the more speculative rail advocates (including me in about ’09) keep saying that ETCS Level 3 will allow rail to do the same, far increasing HSR capacity.

    It’s simple, really. A car going at freeway speeds needs about 4 seconds to brake. This means a stopping distance equal to 2 seconds of driving time. You can get away with closer spacing at lower speeds, and in California people do just that, but any safety margin is based on the assumption that at any moment the vehicle ahead of you could stop instantaneously. It’s called the brick wall rule and it’s not changing.

    jonathan Reply:

    Alon, you have obviously never driven on 101 in the Bay Area during commute hours. Or even weekends.Two seconds driving distance? Ha!

    Alon Levy Reply:

    Guilty as charged.

    But yeah, in California I’ve heard they keep shorter spacing between cars. Can’t tell if it’s because it’s slower or what.

    Richard Mlynarik Reply:

    I think you’ll find that nearly all traffic on nearly all roads follows at unsafe (not within reaction time and braking distance) separation.

    Road traffic entails mass carnage and human suffering due to egregious, basic safety violations that would not be tolerated anywhere else.

    Rail signalling (and building construction codes, and electrical wiring codes, and medical practice, and air traffic, and … and pretty much anything you can think of except US mass marketed weaponry) is held to infinitely higher standards than road traffic, where unskilled, incompetent, dangerous, distracted and hugely fallible amateurs (and fallible professionals) go right out and kill a million and a half human beings a year.

    I don’t, however, support relaxing rail signalling standards to less than “brick wall” braking separation, as costly and as objectively wasteful as that might be. Irrational human psychology means that there’s more at stake, and there is no choice but to embrace this particular double standard even when pretty much every motorist flouts it all the time.

  11. Brad Templeton
    Sep 7th, 2012 at 16:44

    Thanks for the comments. Of course I did not include everything in this short article, but to addres a few issues:

    The question is whether ridership projections for CHSR make sense in the face of future technologies. Those ridership projections govern everything — the cost that must be charged for rides, the efficiency of the trains in energy/passenger-mile, the frequency of the trains that can be sustained (and thus the convenience of the service and its attractiveness.)

    Clearly there is disagreement on the future of cities, and whether it will be polycentric or reverse to greater density and TOD approaches. I believe people will do what they want, in spite of our efforts to legislate. People without children love cities. People with kids change their priorities to revolve around their kids and move to burbs, in spite of hating the burbs. The reality is we don’t know enough to reliably predict which direction this trend will go.

    Peak Oil may be an accepted theory here, and I don’t want to open that argument but instead say this isn’t about oil. Cars will use a variety of power sources, including the same electricity as the trains (though with the cost of battery) and other potential fuels. Robocars allow you to use the right vehicle for the trip — electric short range vehcle in city, longer range electric or liquid fuel intercity, or even a mix of vehicles with 1 minute switchover. Don’t think they are limited in the same way cars are today with fuel choice.

    Such cars will eventually — starting in the late 2020s — allow roads to handle vastly more vehicles than they handle today — as much as 15x as many vehicles in a rush hour commute on the same roads. This does not apply on I-5, however this again was not my point. The CHSR is forecasting that much HSR ridership will come from people currently driving in places like the central valley. I point out that this analysis does not take into account how much better driving (or rather not driving) is going to get in the central valley and other locations.

    People will indeed drive a lot less with robocars — but that’s kinda the point, isn’t it?

    As I write, the HSR will offer a superior trip to a robocar for the trip from downtown SF to downtown LA. But the issue is that this is a small minority of trips. On the trip from Santa Cruz to Malibu, or from Livermore to Pasadena, will it be superior? How about the trip from Tracy to Visalia? The door to door trip, not the downtown to downtown.

    I do indeed like riding on trains and find them much more comfortable — you think people don’t know this? I’ve writing this in France after a nice TGV ride. Enjoying the train doesn’t mean it’s economical, however.

    As for the airports, you know that SJC is seriously underutilized, right? It has no desire to “open up slots.”

    Airlines will innovate in any way that preserves and raises profits. Including finding more efficient planes, other energy sources if oil grows in price. Whatever it takes.

    A downtown station is not quite as superior as you imagine. Getting to downtowns is often the hardest local trip to make for those not on transit lines (and even for those who are.) But again, what matters is not where the station is, it’s the door to door time.

    And while you think that HSR has a smaller carbon footprint, that is in fact incorrect unless it gets astounding ridership. And the whole argument here is that this is a tough sell with all these other technologies competing. No transit system in the world, not even East Japan Rail, uses less energy per passenger mile than ultralight 1-2-person electric vehicles, for example. Those ultralights would not be what goes on I-5, but it’s important not to underestimate the carbon footprint of trains because one must examine the footprint of the entire system, both the full and empty trains, not just the idealized footprint of the rare full train.

    And yes, telepresence robots, seriously. You should look into it, companies to sell them are popping up left and right.

    I believe sprawl comes form more than cheap oil, cheap credit and favourable laws. I believe it comes from parents who use these tools to get what they want for their children. Few want an isolated, drive-to-the-convenience-store life for themselves, though they may wish privacy and space. Mostly they want space and safety and good schools for their kids.

    If they can get it, using cheap electric cars (not sure why you imagine the future of cars is all oil) they will take it. Housing in SF is more expensive than in the burbs, so I am not sure why you think cheap credit is what moves people to the burbs.

    But let’s get back to the central contention. All these technologies, and new ones we have yet to even imagine, make it much harder to generate high ridership projections with high confidence. Without ridership, HSR makes no sense. With high ridership, it makes sense, but only with very high ridership.

  12. D. P. Lubic
    Sep 8th, 2012 at 06:06

    Just stumbled onto this (Christian) perspective on driving which happens to include self-driving cars–from Mars–in 1959! Great jazz soundtrack, too.

Comments are closed.