Elon Musk’s Ideas About Transportation are Boring

Four years ago, I broke my comment section by declaring that Elon Musk’s Hyperloop proposal had no merit, combining technical criticism with expressions like “barf ride” and “loopy.” Since then, Musk seems to have quietly abandoned Hyperloop, while the companies attempting to build the technology, run by more serious people, are doing away with the promise of reducing construction costs to one tenth those of conventional high-speed rail. Instead, Musk has moved to a new shiny target in his quest to sell cars and compete with public transit: The Boring Company. I criticized some of what he was saying in Urbanize.LA last summer, but I’d like to go into more detail here, in light of a new fawning interview in Wired and an ensuing Twitter flamewar with Jarrett Walker. In short, Musk,

a) has little understanding of the drivers of tunneling costs,
b) promises reducing tunneling costs by a factor of 10, a feat that he himself has no chance to achieve, and
c) is unaware that the cost reduction he promises, relative to American construction costs, has already been achieved in a number of countries.

The Boring Company’s Ideas of How to Cut Costs

There is much less technical information available publicly than there was for Hyperloop. However, The Boring Company has an FAQ including an outline of how it aims to cut construction costs:

First, reduce the tunnel diameter. The current standard for a one-lane tunnel is approximately 28 feet. By placing vehicles on a stabilized electric skate, the diameter can be reduced to less than 14 feet. Reducing the diameter in half reduces tunneling costs by 3-4 times. Second, increase the speed of the Tunnel Boring Machine (TBM). TBMs are super slow. A snail is effectively 14 times faster than a soft-soil TBM.  Our goal is to defeat the snail in a race. Ways to increase TBM speed:

  • Increase TBM power. The machine’s power output can be tripled (while coupled with the appropriate upgrades in cooling systems).
  • Continuously tunnel. When building a tunnel, current soft-soil machines tunnel for 50% of the time and erect tunnel support structures the other 50%. This is inefficient. Existing technology can be modified to support continuous tunneling activity.
  • Automate the TBM. While smaller diameter tunneling machines are automated, larger ones currently require multiple human operators. By automating the larger TBMs, both safety and efficiency are increased.
  • Go electric. Current tunnel operations often include diesel locomotives. These can be replaced by electric vehicles.
  • Tunneling R&D. In the United States, there is virtually no investment in tunneling Research and Development (and in many other forms of construction).  Thus, the construction industry is one of the only sectors in our economy that has not improved its productivity in the last 50 years.

This is not the first time that Musk thinks he can save a lot of money by reducing tunnel diameter; he said the same thing in the Hyperloop paper. Unfortunately for him, there is literature on the subject, which directly contradicts what he says. In my Urbanize piece, I mention a study done for the Very Large Hadron Collider, which compares different tunnel diameters across various soil types, on PDF-p. 5. Two tunnel diameters are compared, 4.9 m (16′) and 3.9 (12′). Depending on soil type and tunnel boring machine (TBM) drive, the larger tunnel, with 1/3 larger diameter, costs 15-32% more.

Subsequent pages in the study break down the costs per item. The TBM itself has a cost that scales with cross-sectional area, but is only a small minority of the overall cost. The study assumes five drives per TBM, with the first drive accounting for 75% of the TBM’s capital cost; in the first drive the larger-diameter tunnel is 32% more expensive, since the TBM accounts for 25-40% of total cost depending on diameter and rock, but in subsequent drives the TBM accounts for about 5% of total cost. Another 6% is muck cars (item 2.05, PDF-pp. 7 and 46), whose cost rises less than linearly in tunnel diameter. The rest is dominated by labor and materials that are insensitive to tunnel width, such as interior lighting and cables.

But the actual cost is even less sensitive to tunnel width. The VLHC study only looks at the cost of tunneling itself. In addition, there must be substantial engineering. This is especially true in the places where transportation tunnels are most likely to arise: mountain crossings (for intercity rail), and urban areas (for urban rail and road tunnels). This is why there’s a trend toward bigger tunnels, as a cost saving mechanism: BART’s San Jose extension is studying different tunnel approaches, one with a large-diameter tunnel and one with twin small-diameter tunnels, and the cost turns out to be similar. In Barcelona, the large-diameter TBM actually saved money and reduced disruption in construction.

The Boring Company’s various bullet points after its point about tunnel diameter are irrelevant, too. For example, labor is a substantial portion of TBM costs, but in the VLHC study it’s about one third of the cost in easier rock and 15% in harder rock. There appears to be a lot of union featherbedding in some American cities, but this is a political rather than technological problem; without such featherbedding, labor costs are not onerous.

Tunneling Costs Aren’t Just Boring

At $10 billion for just 2.2 km of new tunnel, East Side Access is the most expensive urban rail tunnel I am aware of. The second most expensive, Second Avenue Subway’s first phase, costs $1.7 billion per km, not much more than a third as much. Is New York really spending $10 billion on just boring 2.2 km of tunnel? Of course not. The 2 km in Manhattan cost a little more than $400 million, per an MTA status report from 2012 (PDF-p. 7). The few hundred meters in Queens actually cost more, in an unnecessary tunnel under a railyard. The cavern under Grand Central cost much more, as do ancillary structures such as ventilation.

The TBM is probably the most technologically advanced portion of urban tunneling today. Even in New York, in the most expensive project ever built, the TBM itself is only responsible for about $200 million per km; more typical costs, cited in a consultant’s report for Rocky Mountain tunneling, are somewhat less than $100 million per km. This is why large-diameter TBMs are so appealing: they increase the cost of the tunneling itself, but save money everywhere else by allowing stations to be constructed within the bore.

Of course, The Boring Company is not building conventional subways. Subways already exist, and Musk likes reinventing everything from the wheel onward. Instead, the plan is to build tunnels carrying cars. This means several things. First, the capacity would be very low, especially at the proposed speed (Musk wants the cars to travel at 200 km/h – excessive speed is another of his hallmarks).

Second and more importantly, instead of having to deal with expensive subway stations, the infrastructure would have to deal with expensive ramps. Musk wants cars to be lowered into the tunnels with elevators. Underground elevators are cheap (vertical TBMs are easy), but in the proposed application they just move the problem of ramps deeper underground: the elevator (“skate” in Musk’s terminology) would carry the cars down, but then they’d need to accelerate from a standstill to line speed, in new tunnels, separate from the mainline tunnels so as to avoid slowing down through-traffic. Trains solve this problem by making the entire train stop in the tunnel and taking the hit to capacity, and compensating by running a long train with many more people than cars could possibly hold. But roads would need the same infrastructure of urban freeways, underground.

Switching between tunnel trunks poses the same problem. Flying junctions are expensive, especially underground. In New York, they were common on the IND subway, built in the late 1920s and 1930s; the IND was expensive for its time, around $150 million per route-km in today’s money, whereas the Dual Contracts from the 1910s and early 20s (with fewer junctions) were about $80 million per underground route-km. Most subway systems don’t do what the IND did, and instead of complex junctions they build independent lines, switching between them using transfer stations. With cars, this solution is impossible, forcing underground four-level interchanges; even above ground, those interchanges cost well into the 9 figures, each.

There is So Much Musk Doesn’t Know

The starting point of The Boring Company is that Los Angeles’s tunnel construction costs, which the company pegs at a billion dollars per mile, need to be reduced by a factor of ten. This means cutting them from about $600 million per km to $60 million. While there is nothing that Musk or his company has said in public that suggests he is capable of reducing construction costs, other parts of the world have substantially done so already.

In my construction costs posts, there are a few projects in the $60 million/km area. Manuel Melis Maynar, the former CEO of Madrid Metro, wrote a brief report on how he built subways cheaply; in today’s money, the underground parts of Madrid’s 1999-2003 subway expansion cost around $70 million per km, but this includes rolling stock, and without it, actual cost is likely to be where Musk wants it to be. Recent subway lines in Seoul have also been in that area, including Metro Line 9 and the Sin-Bundang Line. Going up to $100 million per km, there are more lines in Stockholm.

Melis Maynar’s writeup ignores any of the technological pizzazz Musk thinks of. Instead of trying to squeeze more power out of TBM, he emphasizes good contracting practices, and separation of design and construction. Like Musk, he believes that faster construction is cheaper, but he is aware that the limiting factor is not boring speed: even at a conservative rate of 15 meters per day, a TBM could excavate several kilometers a year, so it’s better instead to begin construction at several points along the line and work in parallel rather than in sequence. Adding TBMs does not make projects substantially more expensive: one TBM used for East Side Access cost $6-8 million, and other estimates I’ve seen only reach into the 8 figures, for multibillion dollar projects. Nor does adding staging areas raise cost underground, where there are many potential sites; underwater it’s a bigger problem, and there costs are indeed much higher, but nothing that Musk does seems designed around underwater tunnels, and his proposed map for LA road tunnels is underground.

Musk’s Ideas: Loopy and Boring

Americans hate being behind. The form of right-wing populism that succeeded in the United States made that explicit: Make America Great Again. Culturally, this exists outside populism as well, for example in Gordon Gekko’s greed is good speech, which begins, “America has become a second-rate power.” In the late 2000s, Americans interested in transportation had to embarrassingly admit that public transit was better in Europe and East Asia, especially in its sexiest form, the high-speed trains. Musk came in and offered something Americans craved: an American way to do better, without having to learn anything about what the Europeans and Asians do. Musk himself is from South Africa, but Americans have always been more tolerant of long-settled immigrants than of foreigners.

In the era of Trump, this kind of nationalism is often characterized as the domain of the uneducated: Trump did the best among non-college-educated whites, and cut into Democratic margins with low-income whites (regardless of education). But software engineers making $120,000 a year in San Francisco or Boston are no less nationalistic – their nationalism just takes a less vulgar form. Among the tech workers themselves, technical discussions are possible; some close-mindedly respond to every criticism with “they also laughed at SpaceX,” others try to engage (e.g. Hyperloop One). But in the tech press, the response is uniformly sycophantic: Musk is a genius, offering salvation to the monolingual American, steeped in the cultural idea of the outside inventor who doesn’t need to know anything about existing technology and can substitute personal intelligence and bravery.

In reality, The Boring Company offers nothing of this sort. It is in the awkward position of being both wrong and unoriginal: unoriginal because its mission of reducing construction costs from American levels has already been achieved, and wrong because its own ideas of how to do so range from trivial to counterproductive. It has good marketing, buoyed by the tech world’s desire to believe that its internal methods and culture can solve every problem, but it has no product to speak of. What it’s selling is not just wrong, but boringly so, without any potential for salvaging its ideas for something more useful.


  1. JJJJ

    I really dont get the big obsession over smaller tunnel diameter being this enormous breakthrough.

    Don’t we already have small tunnels? It’s sort of the trademark for the London Tube.



    Can’t get much tighter than that.

    • Nathanael

      Musk’s supposedly “smaller” tunnels are *standard* London Underground Tube diameter.

      In other words, *Musk never bothered to do his homework*.

      He’s a smart man. But a smart man can make really idiotic mistakes if he doesn’t do his homework. And he didn’t do his homework. I have some remedial assignments for Mr. Musk if he actually wants to improve construction costs. First clue he needs to have is that *surface construction is actually the inefficient part*, second clue is that the mobbed-up nature of the US construction business is the biggest cost escalator, and third, he needs to learn a little something about throughput, capacity, and the conical steel wheel used on trains.

  2. Diego Beghin

    Has Elon Musk even proposed an actual design for his tunnels, with elevators, through-tunnels, service tunnels, flying junctions and all?

    • Alon Levy

      I haven’t seen anything about service tunnels from him. I did see a service map under LA (which would be called crayon if it were rail), which has a trunk and a bunch of branches.

    • adirondacker12800

      There appears to be a lot of union featherbedding in some American cities, but this is a political rather than technological problem; without such featherbedding, labor costs are not onerous.

      ARC cost 600 million dollars. So that the governor could be at the ribbon cutting with a gold painted shovel. Projected to cost 10 billion, to keep the arithmetic simple, call it 12, 600 million is 5 percent of the cost. There would have been ongoing lawsuits. Like the more memorable ones from the Second Avenue Subway. After talking about it for 75 years the Archbishop decided to claim that it would disturb the ambiance blocks away. Or the apartment building that suddenly realized that there would be pedestrians on the sidewalk! Those featherbedding unionized contractors need support staff. Watching glowing computer screens. It not just people with actual tools in their hands that inflate costs.

  3. ant6n

    Very well written.

    (On a more technical tangent, those vertical pushing structures are very neat — they’re a bit like the horizontal pushers proposed for building pedestrian tunnels under underpasses. I wonder whether there’s something similar that could build diagonal access shafts, to make escalator accesses for metros)

    • Alon Levy

      I don’t know of any similar technology for slant bores, but I imagine it can be cobbled up if transit agencies start asking. This is of especial benefit because elevators are low-capacity, and the places with the most demand for large-diameter TBMs are big city CBDs where new stations can expect very high ridership.

      • Michael James

        This is of especial benefit because elevators are low-capacity,

        Not the actual (IMO) future of elevators, brought to practicality by the same German companies (ThyssenKrupp and Siemens) that created the TransRapid maglev. The same technology (which fits Musk’s description of “electric” or “magnetic” sled) can be vertical (elevators), sideways or horizontal (travelators), or any angle you want, and–the important factor–can transition between these modes. I’m predicting they will eventually retrofit the (angled) elevators of the Eiffel Tower with these things to cope with the zillions of visitors every year.

        ThyssenKrupp unveils world’s first cable-free, horizontal-vertical …
        Jun 28, 2017 – ThyssenKrupp unveiled the MULTI system, a cable-free, horizontal-vertical elevator system at an 807 foot high test tower in Rottweil, Germany.

          • Michael James

            No, I doubt that is true. The critical difference is that these things can travel at much faster speeds (and their maglev technology also means they accelerate/decelerate faster too, even as it is smoother and more controlled). You know all the attempts and failures at speeding up travelators? The one installed at Montparnasse was designed for 12 kmph but after too many accidents it was reduced to 9 kmph but still was problematic and the whole thing (ie. two-speed mechanism) was discontinued. These maglev things (essentially sleds that people walk on and stop, holding on to a rail) will/apparently have solved this. Though they could do any speed you wanted, I think they have been operating at 15 kmph which is 5 times the standard travelator at a measley 3 kmph (which is why you and I always try to walk on them but are often thwarted by other lazy travellers).

          • Alon Levy

            Yes, I know about the Montparnasse travelators; give me some credit. And at least in what you linked to it doesn’t seem like a hop-on-hop-off system like an escalator.

          • Michael James

            Yes, of course I know you know about Montparnasse (not least because I have discussed it previously on these pages:-) but the point was to highlight the speeds that are deemed worth aiming for. The ThyssenKrupp technology can seriously outperform those. As I have said before (here), this technology (and cable-free elevators) will begin to revolutionize pedestrian movement, both vertically and horizontally, in the next decade. Think of more of those long travelators at Montparnasse and Chatelet-les-Halles etc. not only feeding into Metro lines/platforms but begin to substitute for them too. I still think they should be considered to link Penn Stn and Grand Central, compared to the billions of dollars in compromised subway line options. In some ways it is similar to some of the things Musk is fantasizing about, but for pedestrians (yes, very un-American). I can see a network of these things in CBDs that links districts, buildings and other transport modes (for longer distance).

      • anonymouse

        Diagonal TBMs are a thing, and Moscow has used them to dig some escalator shafts for deep level stations. There might even be some at work right now. Incidentally, Moscow, one of the most conservative properties when it comes to tunnel designs, is also cautiously experimenting with wider two-track tunnels. There is a line featuring such a section that will be opening in the next year or two, and once they gain some experience, it’s possible that will become a new standard, at least in some cases.

  4. ant6n

    One side note: ironically, the cost of underground flyover junctions is less of a concern for Barcelona’s line 9/10 due to large diameter tunnels, due to the stacked tracks.

  5. Anis

    Great article – and as a non-American, I thought your concluding paragraphs were especially poignant. Thank you, Alon.

  6. Korakys

    To be honest the main reason I RSS to your blog is to read these occasional eviscerations on Elon Musk’s more loopy ideas, or as we would call them here “batshit insane” ideas. There are few places I can go to for this as so much of my usual haunts are in the grip of Musk fever, so thanks. I think the guy has about half brilliant projects (Pay-pal, Tesla, Space X), but also about half really ill-conceived projects (hyperloop, boring, colonising Mars). Capitalism in action I guess.

    • Alon Levy

      The space geeks I know seem to think very highly of Musk. The transportation geeks, not so much. This isn’t too surprising if you look at established companies: Boeing is great at aerospace, but its attempt to make light rail vehicles was an embarrassment. It’s completely fine to be good at one thing but not at other things. It’s just not how American entrepreneurs, tech or otherwise, think of themselves.

      • Korakys

        Yeah, space geeks generally love him and for good reason I think. Unfortunately that means I’m having trouble finding space press willing to stick their necks out to write a take-down of the concept of colonising Mars.

        • Nathanael

          Yeah. Well, I suggest you contact biologists for a really solid takedown. The reasons colonizing Mars is insanely stupid and won’t work are essentially *biology*, not physics. Start with the people who explained why Biosphere 2 didn’t work — they could probably do a really good takedown.

    • Michael James

      I think the guy has about half brilliant projects (Pay-pal, Tesla, Space X), but also about half really ill-conceived projects (hyperloop, boring, colonising Mars).

      The thing is that none—absolutely zilch–of any of those are remotely his “idea”. The concepts are decades, if not a century, old. And Musk didn’t conceive of, invent or code anything to do with Pay-Pal. He was one of several founders of a company scrabbling around for viable ideas (after, I presume, the original idea that led to the founding of the company went down the toilet as is often the case, fair enough) when Pay-Pal fell into their lap. Now, Musk did have the sense to see its potential (some of his co-founders didn’t and left that company because he wanted to bet the company on it and sure enough he was correct). So, fair enough but my main objection is that I can’t see a single thing I would call visionary or revolutionary or particularly “brilliant”. (Pay-Pal is the only contender and Musk doesn’t deserve any of the real credit other than bringing it to market and selling it for billions to eBay).
      But possibly even more irritating is the fawning of the mainstream media. Recently, as he got permission to dig a tunnel under a public street in LA next to one of his premises, it gets reported as Musk testing his revolutionary boring machine. It is nothing of the sort–and to be fair, the company does officially correct that impression–because they are using a second-hand TBM of small diameter that had been used to dig a sewer in Oakland. They have named this midget TBM “Godot”. The tunnel under Hawthorn is to test the “magnetic sled”. Perhaps that is fair enough though it simply sounds like he is reinventing the wheel already perfected by several other companies (German, Japanese, Korean and now Chinese). But actually the real goal may be to have another IPO for another Musk bluesky company so as to recycle the money to keep his various entities going for another few years. Oh, and of course, to harvest government subsidies.
      Curiously, in the article (link below) on his LA tunnelling the journal chose to illustrate it with the two massive TBMs used to build a toll-road tunnel under my city in Australia! (One is blue and one is red, to represent respectively the Jacaranda trees and Poinciana (Delonix regia) trees whose flamboyant blooms colour the city at certain times of the year, the theme being repeated on the massive ventilation towers at both ends of this 6km tunnel.)

      Elon Musk’s Boring Company gets the OK to tunnel under Los Angeles suburb
      Kim Slowey, Aug. 28, 2017

      • Korakys

        Yeah, I originally wrote “ideas”, but just before posting I realised he didn’t actually start Pay-pal or Tesla so I changed it to “projects”. In my brevity I over-exaggerated the difference by selecting binary categories. In truth Tesla has had basically no direct effect, but in terms of spurring other car makers into electric vehicles it has been brilliant (Nissan is the true master here). The Gigafactory might actually turn out to be big news though.

        Space X might still crash and burn because it relies on massively expanding the apatite for even cheap rides into space (satellites still typically cost a ton by themselves), that is why he wants to send ~1000 internet satellites into orbit, to make the market for rockets. I’m mindful of the space bubble in the ’90s. However I still think Space X will succeed long-term and I’m willing to call it brilliant right now.

        It is no news that the mainstream media peddle in idiocy, after all they are there to give people the information they want, not what is true, and the two seldom line up.

        • Michael James

          It’s just that I have a pretty high bar to call anything “brilliant”. Space X re-usable rockets doesn’t meet that bar. I mean the industry has thought about it for decades if not half a century, so zero points for conception. Points to Space X for implementation. Or whoever made whatever breakthrough–which may well have simply been an incremental thing with technology and computing power, and also I don’t know where it came from–it might actually be from gov research labs, developed to a certain point and handed over to industry (as are most hi-tech R&D projects, industry being pretty conservative on what they spend their own money on).
          I’d give a lot more points for the PayPal concept (which we’ve agreed is not Musk) though not quite “brilliant”, and again the main feature is that they were first in a game with winner-take-all as a rule (so points to Musk for acting decisively to develop it).
          And Tesla is way too early to be awarding points to anyone, least of all on any concept which tracks to the 19th century after all. Electric cars are merely a result of incremental advances in various technologies including batteries and their management. I am also on record (on this blog IIRC) as saying that I still think the big players (Japanese, German and maybe Chinese) will achieve the same endpoint at more or less the same time as Tesla, and they already have vastly more resources and market to exploit. I reckon all these companies (Toyota, Benz, BMW, Audi-VW, more than one Chinese) have advanced projects in the works; BMW’s all-electric MiniCooper has been announced for 2019 and while it won’t have the range of Tesla it will be marketed in Europe and Asia where it is a very popular city car and they don’t have the “range anxiety” of those exurban Americans. Half the hipsters who already own Minis will be wanting to trade up a.s.a.p. and it won’t cost nearly as much as a Tesla (and it will be actually available; in 2016 BMW sold almost 400,000 Minis and about 3m in circulation; maybe they’ll even fund an e-Italian Job:-).

          The only points awarded to Musk re Tesla is in convincing his funders to give him so much leeway (and seriously that is impressive). Toyota have already sold one hundred fold more hybrids than Tesla has sold anything. Incidentally it could still be that Toyota buys Tesla; not only did Tesla buy and use Toyota’s Fremont plant to make the Tesla S but part of the deal involved Toyota acquiring Tesla stock. Oh and the Gigafactory is actually a Panasonic outfit funded by and badged Tesla. All of this is fair enough but will only earn points if he starts enough cars to survive and begin to earn some money back.

        • Nathanael

          Tesla did something important, but the context is worth remembering.

          Musk looked at the EV-1 and RAV-4EV and said “Oh, good, the car companies are making a electric cars.”

          Then a few years later he looked at Who Killed The Electric Car, and said “Oh. I guess I have to do it myself.”

          Tesla is doing *the bloody obvious*, using nothing but old technology — the newest tech in a Tesla Model S is the AC induction motors, which have been standard in subway trains since the 1990s.

          The business genius of Musk was to figure out that the would-be competition had *deliberately chosen not to compete with Tesla*, by crushing their electric car programs. This created an massively great business opportunity.

          This is, by the way, why you are *wrong* about the “big players”. They will NOT reach the same endpoint at the same time as Tesla. It’s not because they couldn’t have done so — they have the resources — it’s because of ATTITUDE. They have CHOSEN not to. And it’s pretty much too late for them now. Not a single one of them is actually planning high-volume production of long-range electric automobiles. If they had started when Tesla started (or simply never crushed the EV1!) they would have been way ahead by now. But they chose not to compete. (It’s been called the “Kodak problem” — Kodak invented the digital camera, but refused to switch its business over to digital and was destroyed by all-digital competitors.)

          Hope this helps explain what is going on with Tesla… I realize it’s not obvious.

          • Nathanael

            Note that Musk predicted the Bolt’s sales in 2017 pretty accurately (He said “under 25,000” and they came in about 1500 under that). He said GM was treating it as a regulatory-compliance car and would only produce enough to meet the CARB and CAFE requirements. That’s how he predicted the numbers. He was right.

            This is Tesla’s business advantage. Not technology — attitude. GM could beat Tesla. But GM chooses not to. If they ever decide to get serious, it’ll take them five years — and now Tesla is starting to build up a lead.

          • adirondacker12800

            That’s roughly 25,000 more Bolt’s than Model 3s

      • Oreg

        The price for most overrated techie still belongs to Bill Gates, a cunning business man but neither the impressive software engineer nor the innovator that the public thinks he is for some reason.

        • Michael James

          Bill Gates is smart (smart enough to have seen the potential in personal computing) but mostly ruthless and extremely fixated on success. But no one (except the very ignorant–no one on this site) attributed great technical innovation etc to him. Indeed he was too slow on developments and conservative for his own good. Without his quasi-monopoly he would have been only a footnote in the history of the PC.

          • Alon Levy

            The technological conservatism Microsoft’s gotten infamous for is mostly stuff that happened after he left and put Steve Ballmer in charge. In the 1980s and 90s, Microsoft was really good at being compatible with software that was running on Windows, whereas Apple had an “if it’s not in the docs, we reserve the right to break your program in the next version of Mac OS” mentality.

          • Michael James

            Hah, I think you are trolling. It was after Gates, then Ballmer, retired that MS moved forward (though who really cared by then?) It was under Gates that they missed almost every major advance in the 80s & 90s including WIMPs, desktop publishing and the internet.
            It gave rise to the maxim: “MS’95=Mac’84”.
            Gates was mimicking Henry Ford who for decades refused to change his early model because after all he was still making money hand over fist. Conservatism, complacency and control (because DOS didn’t have competition) and an obsession with eternal backward-compatibility simply meant locking in the past and not moving into the future. Computing–both hardware and software–was advancing at such a pace in those decades that it was inevitable to abandon the past at some point (was it every 5 years or closer to a decade?) to be able to embrace all the advances and the future. Developers may have complained but they were also beneficiaries in the end.

            As it happens I just read today that Apple has returned some $20 billion to those who sell apps on the App Store in the past year. Best year ever; I think they claim some $70bn in total since the App Store was created. That’s got to be an awful lot of happy developers.

  7. Benjamin Turon

    When reading about Musk’s proposed O’Hara project in Chicago, I was struck by how infrequent the service he was proposing, a car every 15 mins when the El gives you a train every 5 minutes, with a lot more capacity for passengers! His maglev cars would be faster, but waiting 15 to 10 minutes would seem to eliminate the advantage of that! When reading the article, I thought if you want a “High-Tech” transportation using a linear motor in a small diameter tunnel… just go with a “Linear Subway” like those in Japan… LOL! Heck, Elon Musk could subcontract the entire project out to Bechtel and Bombardier!

    This was a very good blog post… Make America Great Again… buy Japanese or Canadian… Maybe we could pick up real cheap old cars from London’s tube… complete with Churchill reenactors… LOL! We need more subways in America so we can hide from the North Korean Atomic Blitz… Keep Calm and Carry On! V for Victory!

    • Ethan (@midflinx)

      15 minute frequency is the minimum the city of Chicago is asking for. Probably under the assumption of a traditional higher capacity train-type vehicles. As far as I’ve read, Elon has not said the service frequency of his Loop will be 15 minutes. The frequency he has in mind could be every 10 seconds, or 30 seconds, or 60 seconds.

  8. Clem

    It will be interesting to see what sort of headway Musk envisions, which is what sets throughput capacity.

    Emergency stopping of a train is very different from emergency stopping of a driverless “skate” carrying occupants who are securely seat-belted– although the car itself would also have to be “belted” to the skate! Supposing everything was properly secured, it wouldn’t be crazy for one of these skates to brake at the same deceleration rate as a Toyota Camry, which works out to 6 seconds from 200 km/h. Because the skate is autonomous, we will charitably assume that reaction times are measured in milliseconds for sensing a dangerous condition and acting on it.

    That’s 600 skates per hour, each carrying the standard 1.2 automobile occupants, or a glorious 720 passengers per direction per hour.

    Another crucial issue for headways is route branching, where after the preceding skate clears the branching point the route must be switched, and confirmed to have been securely switched with enough time left for the following skate to emergency brake if the route isn’t properly set (e.g. if switch equipment fails). Suppose that route setting process consumes 4 seconds each time you switch the route, and that every 4th skate takes the diverging route. Now the timing for 4 skates isn’t 4 * 6 = 24 sec, but 3 * 6 + 4 + 6 + 4 = 32 sec.

    That means you’re down to 450 skates per hour, or 540 passengers per direction per hour. On a very good day when everything is running perfectly. So yeah, 200 km/h speeds, but a throughput that effectively rounds to zero when compared to typical freeway or rail capacity figures.

    Yes, SpaceX has done amazing things, but they have most certainly not changed any of the fundamental laws of rocketry. Similarly, Boring will most certainly not change the fundamental laws of fixed guideway transportation.

    • adirondacker12800

      I want to know why the autonomous cars need to be carried on an autonomous sled.

      • Michael James

        Obvious, innit? Not all cars will be Teslas capable of 130 kmph or whatever.

        • adirondacker12800

          My creaky old 21 year old 4 cylinder econo-box will do that. Not uphill but there aren’t any hills underground.

          • Michael James

            Hah, sure but only after 20 minutes with your foot to the floor, preferably downhill, and just before the motor/gearbox blow up!

            Oh, but my bad: just checked and it is 130 mph so >200 km ph.

          • adirondacker12800

            130 km is 80-ish in miles It does 70 mph in fifth without a problem …well itt’s not so econo at that speed.

        • adirondacker12800

          Very reasonable people are predicting the market for internal combustion powered cars will collapse in a few years. When electric cars are cheaper. they are now when one considers total cost of ownership over a few years.

    • Ethan (@midflinx)

      The plan also includes pods holding 8-16 passengers. The ticket price to move a Tesla on a skate could be 8-16 times more expensive than a ticket to ride in the pods. Musk and other rich folks could afford to move their Teslas, but the majority of people in the tunnel might be passengers in the pods.

      Musk tweeted that it hadn’t been decided whether Loop will use rails or tires. If tires are used, track switching equipment isn’t necessary. At up to 600 pods per direction per hour, that’s a capacity of up to 4800-9600 passengers. Very favorable at least to a light rail line.

        • Ethan (@midflinx)

          Not at all. It just wasn’t immediately pertinent to reply to Clem. I’ll copy and edit what I wrote in a comment on a CityLab page:

          Loading 8-16 passenger pods can happen at the curb, just like buses do. Or my guess is for many access points, sites like gas stations will be bought and dug up. With brick-and-mortar retail suffering, more ground-floor sites will be available too, even in some multi-story buildings that can no longer find retail tenants.

          Off-street makes building a row of elevators easier. With four elevators for example, as a loaded pod or Tesla is getting on the elevator at the surface, at least one pod is already descending, a third vehicle is accelerating off its elevator into the tunnel, and a fourth elevator is rising up to the surface. The number of elevators may vary between access points.

          To reduce or prevent backups into the street, cities will be within their rights to issue fines if that happens, or require surge pricing discouraging too many cars from queuing. The result is taking a car in a Loop during peak times will be an expensive luxury that subsidizes Loop construction and operation. Many of the vehicles will instead be pods with passengers.

          • adirondacker12800

            That’s not how bus stops work. Why does the whole bus have to trundle up and down the elevator at every stop?

  9. Archit Goel

    LOL…I would have hoped you could examine both the sides of the argument, that would have made your article more credible. Anybody who can only argue one side is inherently biased. You can use selective facts to make any case.
    Musk’s cult members (I am one) like to talk about Space X and Tesla. Because the arguments about technical feasibility, practical considerations, etc were much higher there.

    What you forget is a very simple point. About what people believe in the cult of “Musk”.

    its all about the vision. Your normal companies have “puny” visions of increasing shareholder wealth, make efficiency improvements, building a better widget. Which is NOT wrong, but it’s not inspiring. Its OK, not Great.

    Musk’s vision is simply stupid, audacious and civilization changing. If Boring company’s vision of “unlimited no of 3D tunnels” is realized, the urban design will be forever changed.

    Of course, their are all manner of practical challenges. But Musk will not “agree” to current “wisdom” and “limits” agreed by industry. For him, only limits are ones imposed by Physics. Reset everything can be worked around.

    This is why Musk will probably succeed. Because no one tries to do the impossible, so they cannot do the impossible. Only if you deliberately try to do the impossible, try to IMPOSE your ideal reality into the world, can you change the world.

    • Eric

      “LOL…I would have hoped you could examine both the sides of the argument, that would have made your article more credible.”

      Would it be “more credible” to examine “both sides” of whether the world is flat? Whether insects are spontaneously generated? Sometimes there is simply no evidence for a proposition, and you just look stupid for arguing it.

    • adirondacker12800

      Henry Ford’s wife drove an electric car. jay Leno has a similar one in his collection. With what he things may be original equipment batteries. We’ve been waiting for cheaper denser batteries since. It’s why Tesla cars, until recently, used commodity lithium-ion cells to power the car. The rest of it is minor tweaks to already existing technology with a thin veneer of kewl software.

      • Untangled

        Aren’t Tesla batteries from Panasonic? Even the gigafactory batteries are using Panasonic tech. So even Musk’s batteries aren’t original.

  10. Jacob

    How about getting gondolas to go really fast? Such as 40-60 km/h. The fastest gondola on the planet is 20 km/h.

    A 40 km/h gondola would solve a lot of problems. For example, you could go from MEL airport to the CBD in 30 mins. A heavy railway is not under construction because of concerns regarding the operating costs of the trains.

    • Michael James

      You might be talking about:

      Or maybe Trackless Trains (aka “Tubular Rail”):
      Long rigid train spans two pylons without anything else supporting it. The train has rails attached to it which are held by wheels (rollers) on the pylons and the trains slides between the pylons.
      (I don’t have the web address but this was about 2009):
      The ‘hole’ skinny on Tubular Rail
      August 10, 12:25 PM · Alan Kandel – Fresno Green Transportation Examiner:

      Called stanchions, the proposed supports (see illustrations) house the rollers with the trains’ rails passing over (or rolling on) the rollers; the rollers themselves to be motorized. The stanchions, spaced accordingly, provide what’s called cantilever support, with no fewer than three stanchions supporting a train at any one time – sometimes four depending on a train’s relative position. This way, support integrity is always maintained.
      Each train would be capable of holding 400 to 500 passengers running at speeds of up to a recommended 150 mph for fast travel applications. These systems could also operate slower-speed in urban settings. Since the idea is to move the trains above the ground surface, there is no interference from either intersecting vehicle or pedestrian traffic. Because of the propulsion technology used, there would be virtually no emissions released into the air. Expect infrastructure costs to be lower than those of high-speed systems, estimates being about $20 million per mile to build, according to information on the website.

  11. orulz

    I agree that from what they’ve focused on showing of their proposals so far there’s not much value.
    But I am also not quite so pessimistic as to say that there’s absolutely nothing there to salvage.

    1. Tunnels that only (or primarily) carry cars on skates are a stupid, stupid idea, full stop.
    2. You have to remember, though, that 16-passenger ‘people pods’ are big part of the proposed system, too. You can get respectable (but not earth-shattering) throughput with 16-passenger pods operating at highway-like headways You could think of it as “Jitney Rapid Transit.”
    3. Using emerging autonomous vehicle technology in place of conventional signaling systems could save money.
    4. Smaller diameter tunnels ARE definitely cheaper; perhaps not as cheap as Musk claims, but cheaper nonetheless.
    5. If they can’t come up with a usable, convenient, and inexpensive way for this system to interface with ground level, then there’s no hope.

    • Alon Levy

      Ad 4, my claim isn’t “costs are insensitive to tunnel diameter”; it’s “costs grow sublinearly in tunnel diameter, so costs per unit of capacity are lower for larger-diameter tunnel.”

  12. The Economist

    Apart from the fact that there are too many stakeholders (unions, politicians, NIMBYs) that need to be paid off, there is only one other way to cut costs: eliminate the non-essential “junk” construction elements present in subway or rail tunnel construction. That is one place where Musk (or someone else) might be able to make progress to reduce costs. As Alon has made the point many times, the cost of the tunnel itself is not the major cost of those projects. It is in the stations and everything associated with them. If Musk can figure out how to build cheap austere but functional stations that still pass the government safety then he will have something to brag about.

    Highways are cheap exactly for that reason: there are no stations on them. The highway “station” is your garage/driveway. Highways outsource the “station” building and maintenance to the average citizen. They also outsource the “train yard” to the same place. You want to build a subway, you need to build a train yard somewhere to lay over and maintain the trains. You want to build a highway, the car parking is not your problem — it is your user’s problem. Additionally for highways, the vehicle ownership and maintenance is outsourced. The highway builder/operator does not buy or maintain the vehicles. If you could build a subway with just tunnels and let the users build their own stations and yards, purchase their own vehicles and maintain them, then the cost of the subway would not be bad at all. Those are the things where private enterprise might be able to come up with some implementation that is much more cost efficient than the usual way subways are built and run by governments. I agree with Alon, the idea that one will make subway construction much cheaper by only making the “boring” process more efficient is bogus.

    Highways by their nature are very austere: asphalt, concrete and steel. Nobody wants to live close to them. Nobody wants to walk along them. Nobody wants to hear the noise they generate or breathe the pollution they create. A pure subway tunnel is very similar in nature, to the extent that is made of concrete and steel and the average person does not want to be inside it or walk along it. The difficulty is that is the stations are pure concrete and steel, even when perfectly clean they still would not be attractive to the average person. The average person will only go to such a station if the benefit outweighs that repulsion. Compare that to the inside of a car — most people find it much more attractive with its “personal” space separating you from the rest of the users. [That is indeed why on some services in Europe the cars are split into cabins — so that the space is shared with only a few other passengers.]

    Don’t get me wrong. I am not advocating for more highways. I am just pointing out the natural disadvantage that subways have relative to the highways. Working to alleviate those is likely to save much more money than whatever can be saved with better TBMs.

    • Joey

      You’re comparing above-ground highways with underground trains. For above-ground trains, the bulk of the cost comes from the same place – structures.

      The reason that underground stations are so expensive is that you have to dig station caverns (at least for certain construction methods). Once that’s built whatever goes inside it is relatively cheap.

      So I don’t think this argument really holds up.

      • Michael James

        I agree that Economist’s (!) argument doesn’t hold up, but then I don’t think his econocratic thinking holds up either. As Alon points out, European or Asian metro projects cost a lot less than American ones, and the former often take great care in the design and aesthetics of the stations. Anyway the economics is false. Building a metro line or system is building for the ages and it is entirely wrong-headed to skimp on a bit of aesthetics especially when millions of people will use it every day. The three originals (big city) are more than a century old and by comparison with the other two (London, Paris) the NYC system may be functional but it really is a depressing and even oppressive place to be. Even when it is clean and perfectly safe it looks shabby and has the appearance of a neighbourhood you wouldn’t walk into if you could avoid (and I’m not talking the 70s period when it really was something to avoid).
        Here is an extract from a piece in today’s CityLab on the opening of the Toronto Spandina line:

        The Ambitious Design and Low Density of Toronto’s Newest Subway Stations
        CHRIS BATEMAN DEC 18, 2017

        Despite its shortcomings, the scope of the 5.3-mile Spadina line extension is ambitious. Each of the six new stations is a unique creative expression that blurs the line between public art and architecture; a philosophical continuation of the original Spadina line that opened in 1978, and was the first Toronto subway line to feature stations designed by multiple architects.
        “We’ve found that when you do good urban design, good architecture, the local communities take ownership of the stations,” said Ian Trites, supervisor of architectural design for the Spadina line extension.
        “Kids are less likely to vandalize them, there’s lower maintenance as a result, and people actually enjoy being in the stations,” he said.
        The TTC asked local architects to partner with international firms on their designs, a move that resulted in heavy-hitters like Will Alsop, Foster + Partners, Aedas, and Grimshaw Architects shaping the design of the Spadina line extension stations.

        • Alon Levy

          The station design in New York isn’t bad – the station mosaics are cute. The problem is poor maintenance. The well-maintained subway stations in New York don’t look any worse than the Metro stations here.

          • Michael James

            I have to admit I haven’t used the NYC subway very extensively but my overwhelming memory is of the subterranean stations that don’t have ceilings and have rust-tinged steel I-beams, pipes and wires etc with bare fluoro lighting. The passageways are often too small and grotty too. The above-ground stations are not much different to Metros elsewhere, perhaps including Paris. But I don’t find any of the Paris Metro stations oppressive–almost all of them have the big arched ceilings, most with white tiles that at least make it light and clean looking, combined with the neat large-scale wall advertising (that provides a bit of diversion while waiting for a train). I think London looks and feels a bit grotty by comparison to Paris but still different league to NYC.

            Incidentally while there are books on the NYC subway, there are actual art books and photography books on the Paris Metro. I have the Joseph Giovannini book on the NYC subway (Subway Style) and it is notable that it hardly has any wideshot pics of stations, instead having closeups of various bits of it: seats, ceramic station signs, metal work etc. Not like the ones on the Paris Metro. I’m just leafing thru the large-format one (Metro: Photographic Elevations of Selected Paris Metro Stations) by Larry Yust in which he took a series of photos while walking the length of the station then stitched it into one pic with software. I find it peculiarly evocative compared to regular photos (possibly because it does mimic the way we perceive it and remember it?). I also have the one by Actes Sud featuring “the major figures of photography all snapped the Paris metro, from the humanists–Doisneau, Henri Cartier-Bresson, Brassaï, Boubat, Izis, Kollar, Ronis and more–to photojournalists such as Robert Capa, William Klein and Van der Keuken, in addition to the scores of photojournalists who passed through the city.” Perhaps Moscow and maybe Stockholm are more elaborate but I don’t think any other system has been so documented by so many, as Paris has. There are reasons for this. It’s actually one of the big things that makes me homesick for Paris! Give it time Alon, it’ll grow on you.

          • Alon Levy

            “Small and grotty” describes the passageways here pretty well, too, at the big transfer stations. I basically gave up on Chatelet and when I take M1 to that area I hop off at Hotel de Ville.

            Stockholm is completely different from Paris. The rock is very hard and forms natural arches, so the tunnels have no lining, and the stations show bare rock. There are also some sculpture exhibits at the stations, which there aren’t here or in New York (here they instead have plaques about the city’s history).

          • Michael James

            Looking thru another of my books (New York Underground, Anatomy of a City by Julia Solis, 2005) Fig. 7.1 is a drawing: “Alfred Beach proposed a pneumatic underground railway running from the Battery to Harlem”. This was 1867-1870 (in response to London’s pneumatic tube train that ran in 1863). Alfred was the Elon Musk of his day! His pilot tunnel under City Hall remained until 1912 when it was replaced with the first real subway:

            Figs 8.4, 8.5 and 9.2 are of “City Hall station, closed since 1945, still retains some of its former splendor”.
            These have tiled and decorated vaulted ceilings by Rafael Guastavino. I guess that style was abandoned early. Same thing might have happened to Paris, ie. moved to more unadorned naked engineering, but they built most of the system in a blitz in those early years. I’m grateful.

          • johndmuller

            The DC Metro is worth talking about in the design context. The original designs were basically “not like NYC” – which they translated into airy open stations with long sight lines and minimum amounts of unsightly signage, tawdry commercialism or unsurveillable corners. The aesthetic was clean, toned down, cool and quiet, and it kind of worked, definitely as far as being not NYC.

            Aside from the austere and consistent look, there’s nothing really eye popping, although some of the main transfer stations, like L’enfant Plaza and Metro Center are pretty impressive just as huge underground caverns. Clean is impressive too, can’t deny that clean is a big subway deal just in itself. Not so much as a work of art, except maybe insofar as the Union Station stop is paying homage to its definitely artsy big brother above, you come up out of the one vault and there you are in the middle of a much bigger and more ornate vault.

            Gradually though, the visual clutter has made inroads – bigger signs, more lights, more electronics, ads, etc. – as yet still pretty clean. Unfortunately, the train operations themselves have been going downhill and service is said to be less than desirable.

  13. Pingback: News Roundup
  14. Tunnel Vision

    Most f what Musk is proposing is already being done by TBM manufacturers and the like. There is a limit to how fast these machines can go depending on the ground conditions. If they can be more automated that would be a start but until he actually designs and manufacturers a TBM that can cope with all the varying ground conditions from Granite through running sand which require different technologies then judgment has to be reserved.

    One comment I want to take issue with is your comment about the unnecessary tunnels in Queens for the ESA Project. Given that trains cannot levitate just how do you propose to get the new revenue tracks from the existing 63rd St Tunnel elevation to link in with the existing revenue tracks into Penn Station. The unnecessary tunnels that were built are pretty much at maximum gradient for third rail trains as it is. You cannot realistically be proposing that these connections should have been constructed using open cut techniques which would have caused immense disruption to AMTRAKS Sunnyside Yard and Harold Interlocking.

    As for ESA construction costs some 38,000 l.ft of new tunnel was constructed together with the caverns in Manhattan.
    The excavation costs for those tunnels, cross passages, 5 caverns, 4 vertical and 4 escalator shafts and 6 junction structures has totaled around $1.2bn. The lining and fit out has cost another $1.3bn. More than in other Cities for sure but tell me how many projects like ESA have been built where your main access for men, equipment and materials is 3 miles from where you are actually constructing the work, access in Queens at the so called 63rd St Bellmouth and construction in Manhattan.

    I’ve read a lot about construction costs in NY and other Cities and while much of the criticism of NY related to Union over staffing, politicized contractors may be true and contribute to increased costs there are other factors at play for specific projects which receive less attention. For example, on ESA the work in Harold Interlocking has long been a source of major problems leading to cost and schedule overruns. The interplay between AMTRAK and LIRR in that location has resulted in territorial issues where you can end up with 8 railroad employees watching (protecting) a four man ironworker team install a catenary pole. Also getting support for 3rd party contractor work in this are has been archaic to say the least. Requests for support from the railroads are submitted 2 weeks ahead of the panned work, they are reviewed and the railroads determine what they might be able to support based on their work in the area, then the “jobs” are posted on the work board and the Amtrak and LIRR support staff bid the jobs. You then find out one day ahead of time whether your work will be supported or not. Sometimes it is and sometimes its not. Then come Friday 10pm and guess what the A-man does not show up as he’s sick and there’s no replacement leading to cancellation of your weekend work. Now when your moving signal, comms and power infrastructure there is a sequence that has to be followed and once you lose a critical weekend you might not get back in to that area for a while, plus now the MTA has to pay the Contractor for the crews that turned up but did not work and the knock on delays to their critical path work. One of the Contracts on ESA was delayed so much that the delay costs equaled the original value of the Contract……could this have been avoided, well of course but there’s little to no political will even on the FTA’s behalf to take on the relevant parties.

    Another major driver is the way projects are funded through the 5 year MTA Capital Plan. Its not uncommon to receive approval of the 5 year plan in Year 2 of the period that is covered by the plan and that the approval is for less than requested meaning decisions have to be made with regard to which contracts are funded and which aren’t. When Capital projects schedules are beholden to the political will of Albany its no wonder they suffer as the MTA is an easy target.

  15. Pingback: Catbus» Blog Archive » Far From Boring:Meet the Most Interesting Tunnel Boring Machines
  16. Pingback: Elon Musk's "Innovations" are Not the Future — They're Delaying It | Copy Paste Programmers
  17. Pingback: City Announces Competitors for O’Hare Express, Including Noted Transit Hater Elon Musk – Streetsblog Chicago
  18. Pingback: Elon Musk Não é O Futuro | O Minhocário
  19. Dev

    Apropos of nothing, I find this website incredibly hard to read – thin Helvetica typeface coupled with small default point size and crowded CSS line-height conspire to drive this reader away.

  20. Pingback: Tesla, Elon Musk e a crença miserável na tecnologia
  21. Pingback: Gelecek Elon Musk ile gelmeyecek • vesaire
  22. Pingback: [Перевод] Илон Маск – это не будущее — iT ОБОЗРЕВАТЕЛЬ
  23. Pingback: Elon Musk não é o futuro – Jacobin Brasil

Leave a Reply

Fill in your details below or click an icon to log in:

WordPress.com Logo

You are commenting using your WordPress.com account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )

Connecting to %s

This site uses Akismet to reduce spam. Learn how your comment data is processed.