Hacker News Reader

Comment by comebhack | Hacker News Reader

There is some more detail on the bridge itself in this video: https://www.youtube.com/watch?v=8tpv6n1ykfA

The bridge is assembled over 2 nights at a motorway exit (so traffic can bypass it by driving off and immediately back on to the road). During night 1 the two end ramps are assembled and attached together to make a short bridge. During night 2 the ramps are driven apart, the central section is built to reach the full length and the entire structure is driven to the final location.

The entire length is 236 meters long providing a working length of 100 meters underneath. The assembled bridge can flex slightly at the joins between sections, and has a turning radius of 2 kilometers.

View on HN

The whole Marti youtube channel is a marvel for engineering geeks like me. If you have the occasion, you should take a look (talks about tunneling, big machines, etc)

This one https://www.youtube.com/watch?v=6AV2NcyX7pk is insane. Elon's Boring Company is a joke in comparison.

They use a tunnel boring machine to bore a tunnel with a 45° slope.

They do go into the mechanics of how they make this insanely massive machine drive up a grade that steep, and how they ensure it doesn't slide backward.

I was glued to the screen more than with most movies.

If you like channels like Practical Engineering, you will enjoy this.

I watched this at 2x expecting a mildly interesting educational video. This was a mistake. Almost forgot to blink.

I was glued to the screen more than with most movies

i was going to say something similar. this was better than the action movie i saw earlier.

Agreed - most interesting YouTube video I’ve seen in many months.

Elon's company is targeting different things - cheap and fast, unlike the company in the video doing technically hard but probably not cheap.

Is The Boring Company actually targetting anything?

Or was it just a silly idea that Musk threw some money at and they've been struggling to justify their existence ever since?

Their website has:

>The mission: solve traffic, enable rapid point-to-point transportation and transform cities

via cheap tunnels. Maybe they won't get there but it's an interesting goal.

Move fast and break things runs into problems when “break things” translates to collapsed tunnels.

The boring company isn't move fast and break things. They literally just bought a drill and use it like any other construction company. There's no innovation at all.

They started that way, just buying a drill, but are now trying to innovate and build their own one. They are just starting using Prufrock-3. The Prufrocks are machines the Boring company have made themselves. They say "Prufrock's medium-term goal is to exceed 1/10 of human walking speed, which is 7 miles per day." which is way faster than anyone else. The Swiss machine in the video did 400m in 4 months. They are also experimenting with evacuated tunnels (https://youtu.be/nV07jqwCy0A?t=879) which may not go anywhere but is at least an attempt at innovation.

"7 miles per day" is all talk and no proof it can actually be done. The Swiss dig a lot of holes from easy to very complex terrain. If there was an easy way go that much faster they would.

There was a derailment of a cargo train the the Gotthard base tunnel last year [1]. It caused a huge amount of damage and one tunnel will be close until September this year. They need to replace 7km of track. If there way anyone to do this correctly faster they would as it is costing millions not having this route open. Concrete needs time to cure etc. some things you can't just make faster because someone said so.

[1] https://www.swissinfo.ch/eng/business/gotthard-base-tunnel-t...

It’s incredible how defensive people get around others’ ambitious goals.

Every Elon company sets super high goal targets. They are often completely unrealistic with current technology but it inspires certain types of people to innovate and it works quite well.

> The Swiss dig a lot of holes from easy to very complex terrain. If there was an easy way go that much faster they would.

Nobody said anything about “easy”. Also, The argument of “the established players don’t work on X so X isn’t possible” doesn’t work.

If the Swiss are happy with their industry the industry isn’t going to risk really capital intense experimentation to do better. See: Innovator’s dilemma

> ... medium-term goal is to exceed ...

According to the roadmap, another Elon company should have attended fully automated driving years ago.

    > The Swiss machine in the
    > video did 400m in 4 months.

No, a lot of the time the machine was stationary because the needed to manually reinforce what they were about to drill through, so the machine wouldn't get trapped and buried in gravel. Wouldn't Prufrock-3 be similarly slowed down!

In addition to that I assume that the 7 miles a day claim assumes 24/7 drilling, whereas I wouldn't be surprised if the swiss were doing 8/5 drilling.

Well the innovation appears to be, drill smaller holes faster, and put less traffic through them.

Brilliant!

Nobody said “break things”. Making one specific thing reliably, quickly and cheap is a completely different approach to the “move fast and break things” approach.

There's no collapsed tunnels. It's the same technology, just a smaller surface area / volume of displacement due to not needing exhaust mitigation.

Is fast and cheap always better though.

Fast. Fast always wins. Trying something, regardless of outcome teaches something about the world. The more trials, the more you roll the dice, the more you learn.

I really hate being wrong, but it is much better to be wrong a lot, and quickly understand why. The alternative is to try nothing. It’s kinda sad.

Fast wins for something like mobile apps but not infrastructure where safety matters and you can’t just shrug off liability. The Boring Company is a great example: fast to market themselves but almost all of their projects have fallen through.

Which nascent technologies have became common, dominant by innovators unilaterally adhering to the precautionary principle?

The cliché "Worse is better" describes this (turrible) phenomenon. (The flip side being "safety regulations are written in blood.")

I really wish this wasn't the world we live in. I want to live in a world with consequences (and justice). I've been railing against it my whole career. And judging from my meager savings, failing in my efforts.

Iteration always wins. There may be other constraining factors, but if you can violate those to iterate more (i.e. cheat), you’ll beat your competitors who honor those constraints.

This has nothing to do with the success or failure of any given company.

> but if you can violate those to iterate more (i.e. cheat), you’ll beat your competitors who honor those constraints.

Until people die. Then suddenly, you can't cheat any more. And everybody else has to live with your decisions.

https://en.wikipedia.org/wiki/Morton_Thiokol

I wouldn't call that winning.

You seem to have misunderstood what I’m saying. Iteration is how you learn. Learning through iteration, which is known under various names like Wright’s law or more generally experience curve effects [1], goes hand in hand with economies of scale in driving down costs.

Yes, iteration requires you to survive. Not sure how that’s relevant. Cutting corners also isn’t necessarily a bad thing - you’re focusing on the extreme example where people die. Cutting corners can also be a careful evaluation of what processes are and aren’t relevant to a given situation but that becomes trickier when it’s enshrined in law. Imagine if guidelines from the 80s about how to write software were enshrined in law.

Regulations more often than not do ignore the flip side in terms of the cost of compliance because it’s difficult to show the counter factual universe in which a regulation may save 10% more lives (or maybe even 0% more lives) but drove up costs by 100x.

[1] https://en.wikipedia.org/wiki/Experience_curve_effects

> Imagine if guidelines from the 80s about how to write software were enshrined in law.

If Edsger W. Dijkstra's wild ravings from the 80s were enshrined in law, perhaps we wouldn't have quite so many fatalities attributable to crap software. https://www.cs.utexas.edu/users/EWD/transcriptions/EWD10xx/E...

> A number of these phenomena have been bundled under the name "Software Engineering". As economics is known as "The Miserable Science", software engineering should be known as "The Doomed Discipline", doomed because it cannot even approach its goal since its goal is self-contradictory. Software engineering, of course, presents itself as another worthy cause, but that is eyewash: if you carefully read its literature and analyse what its devotees actually do, you will discover that software engineering has accepted as its charter "How to program if you cannot.".

It does depend on the domain though. Sometimes moving slowly and carefully can result in a faster successful outcome than throwing shit at a wall and seeing what sticks.

> The more trials, the more you roll the dice, the more you learn.

If you have infinite dice rolls this is obviously true. If every dice roll costs you something or indeed everything... heh. Maybe don't just roll it to see what happens?

Yes, sure fast wins when building a bridge. Or a tunnel. Which then collapses. Safe wins, this is not Facebook where people share some holiday pictures.

"but it is much better to be wrong a lot"

I disagree, "Oh the bridge collapesed, I was wrong! But this is much better than being right" - Nope.

If you're opposed to fast and cheap, I have great news for you about the cost and pace of construction in the US.

... which has its own reasons, namely regulations (you know, every time there was a new disaster they added another one) or maybe some corruption. You don't see bridges falling in US regularly killing tons of folks, do you.

I am not saying its ideal and there is no room for improvement, nothing in real world is, but please consider other, 'fast' scenarios for long term (100+ years) existence when not only many lives are at stake.

> You don't see bridges falling in US regularly killing tons of folks, do you.

Falling yes. Killing is hit-or miss. I would be surprised if the reasons bridges are failing due to lack of maintenance is entirely unrelated to the cost of building.

There are many ways things could be better, and many ways things could be worse.

Attacking something only because it's cheaper and faster seems silly.

Please share a single recent example from US, I am only aware of one relatively recent catastrophic case from northern Italy, where maintenance was subpar.

Cheaper reflects many things, overall quality testing, attention to detail, how much effort went into design etc. Its not a silver bullet but oh boy does it always show on result, without exception.

https://en.wikipedia.org/wiki/List_of_bridge_failures

https://en.wikipedia.org/wiki/Francis_Scott_Key_Bridge_colla... was not 2 months ago.

> Its not a silver bullet but oh boy does it always show on result, without exception.

Permit me to show you an exception!

This watch is USD59: https://www.target.com/p/men-39-s-timex-expedition-scout-wat...

This watch is GBP225: https://mrjoneswatches.com/products/a-perfectly-useless-even...

There are even more extreme examples, e.g. the watch from Richistan.

Depends on the application, but commodifying (easier instances of) things that previously required bespoke engineering is a legitimate pursuit.

Fascinating. Thanks for that link.

For a contracting company the videos are exceptionally well produced.

They remind me of something more like an early-era discovery channel show like extreme engineering.

They just need to get mike rowe to narrate.

I was more impressed by the OP seemingly from Swiss public body. The Marti one in GP is exactly the sort of sales demo type video I'd expect 'for a contracting company', though with lashings of high speed chase or shooting narrated video (you know, the 40% ads, 50% rehashing what we've seen or telling us what's to come, 10% content variety) for some reason.

For the bridge only, Marti also made a more sufferable cut in German, that is on the OP channel (https://www.youtube.com/watch?v=VJLX3C0eg3g)

The segue from CG is much more sensible

I've watched these. They're excellent.

They make me wonder why these European companies don't compete in US infrastructure projects?

Because there are American (or Canadian, British, etc) companies that do what's necessary for American Civil Engineering.

When it is worth it, American public works entities do reach out to German (and other national) companies. As to issues like this specifically? Because the US generally has wider freeways/highways, so is less impacted by single/double lane shutdowns for surfacing. In addition, many states have opted for less long-lasting quick pack asphalt for surface streets which can be resurfaced in place and ready to drive on again in a few hours.

I know it's Internet rhetoric to assume America and it's government are incompetent, but the Civil Corps of Engineers, CalTrans, etc are actually pretty good at their jobs. The biggest horror stories are jobs given to private entities that go overbudget and overtime.

Arguably a big chunk of issue is that a lot of projects that were built by government now are only paid for by government... through the nose.

One factor you’ll see in many areas of government is the second order cost of eliminating civil service positions. It’s most common to talk about how contractors usually end up costing more and being less efficient due to additional overhead and conflicts of interest, but there’s a deeper problem that the government doesn’t have a staff of people with the knowledge and experience to select and manage contractors. That’s how you end up in situations where none of the alternatives are better than eating the cost of a bad plan or accepting a lower project lifetime, and because it’s a managerial failure the blame is often spread between three or more organizations and often has no effective accountability.

Well, for the California example, CalTrans is the govt. They're well staffed and don't (generally, it may occur occasionally) rely on contractors for their work. I believe the same goes for CDOT and other agencies, though many of them are more willing to contract out as they don't necessarily prioritize roadwork like Californians do (for obvious reasons).

Generally, it's particularly rare (at least in the Western "blue" States) to rely on mostly or, especially, exclusively private public works programs.

It's a monopsony, no? So if you ingratiate yourself with the buyer, you're in.

I think there are some sort of "buy American" restrictions for many US infrastructure projects.

Competition is ruthless.

Try to open a lemonade stand in other kids neighborhood they will kick you out :)

Now try to do the same where millions or more $ are at stake, good luck with that :)

> and has a turning radius

!!!!

There's a mindbending sentence.

Not all roads are straight.

But often, temporary bridges that can be moved and assembled and hold traffic are straight.

Wait so all this lets them pave a 100 meters a week (assemble bridge on one weekend, pave on Monday, then disassemble the bridge the next weekend after the asphalt has dried)? That seems horribly slow and expensive.

100m at 1 lane is around 430SY. That's probably 2hrs of milling and an hour of tacking and paving, with maybe another hour or so for incidentals. So you may only get half a workday of production. For time consuming repairs, like full-depth replacement, the setup time cost may not be significant.

Keep in mind, though, you don't lose a lane of traffic. There is no need to truck in jersey barriers. You don't have to build an entire temporary detour road. You don't pay a consultant $200/hr to design a traffic control plan.

I think the real value is safety. The crew is shielded by the bridge and you have complete grade separation from traffic. That's a lot better than an orange barrel being the only thing between you and a minivan.

430SY? What's SY? Square yards?

Yes. We don't use metric in the USA and you can't make us :)

I'm not American, but I would have understood what you meant had you written sq yd instead.

I have a addon that autoconverts imperial units. Your comment is the only thing peaking out of the insanity bubble. =)

>Yes. We don't use metric in the USA and you can't make us :)

NASA demonstrates otherwise.

And the military, and for science, some parts of government. And wine bottles and soft drinks, but not milk or beer.

It’s like Windows with its half-hearted efforts to modernise it’s interface.

> no need to truck in jersey barriers.

The video clearly shows barriers between the workers and the active surface lane.

Once installed, the bridge can be driven along the road when a 100m segment is finished

oh this is cool. I did think for a 2 day setup and 1/2 day takedown it wasn't a huge efficiency saving but it is if you move it down the road at the same time. As the comment above mentions, safety is a huge factor too.

Presumably the road has to be closed for the bridge to travel? I assume this takes place at night.

During the night, there will be 1 lane open in each direction (one on the side of the bridge, and one on the opposite carriageway), so the bridge can be moved.

If it's moved slowly enough, I don't see why traffic couldn't drive on it.

I guess it's theoretically possible to engineer a bridge that can move with traffic on it.

But this bridge is engineered with solid feet for taking traffic loads. The wheels are only extended for movement and wouldn't be able to take the load of traffic.

That probably just takes a few hours, easy to do. And you can start/stop a few times to let cars go over.

I think they pave 100 m, wait til it's dried (next day, perhaps) then drive the bridge 100 m up the newly paved bit and start again. No break

No. This bridge travels. Look closely: it has wheels and it's motorized.

> That seems horribly slow and expensive.

That’s because it is. Welcome to the world (in the case of USA, the entire country) car centric transportation.

Having to stop traffic, and then redirect it into the one emergency lane, every time 100m is finished in order to advance seems like a huge disadvantage.

If the road is anywhere close to max capacity this will cause traffic jams either way.

But usually roads aren't even close to max capacity at night, when the shifting happens – which, I imagine, is much less stressful and time-critical than doing the whole resurfacing in a single night.

It would be awesome if the the entire bridge could slowly move as one while traffic keeps flowing over it. That would require far more and far bulkier wheels than the current ones designed to carry only one support segment. That will have to remain the stuff of fantasies...

The bridge could temporarily lift just the 2 ends and traffic could continue slowly under the bridge while the bridge moves ahead. However, it needs to also raise its height for trucks to pass under or alternatively, trucks could be temporarily suspended/rerouted from the road while the bridge moves.

If you look carefully, it seems that it can. It has wheels, and it's probably motorized.

It had wheels and it can move. But only with traffic shut down.

Sure, but that can happen at night with minimal disruption to traffic as it takes only the time to move it, not the time to disassemble, move, and re-assemble.

Traffic is currently redirected into the emergency lane 100% of the time, so this is still an improvement.

hence why the work requiring the stopping and redirecting happens at night.

They can only complete 100m once every 12 hours?

with minimal road disruption, though.

at least where I live you'll often seen lane reductions measured in weeks for repaving.

I've seen projects to pave only a few miles take years. Not even roads that are that busy, even.

That doesn't sound right, assuming a lane width of 3.5m, they can only resurface roughly 29 square meters an hours?

With all that equipment and manpower.

you need to scrape all the way to the base and then fill back in when repaving. and you also can't lay asphalt in rain.

I would think that merely repositioning can be done overnight.

I was under the impression the bridge rolled forward as the works continue.

Possibly, but unlikely IMHO - it looks like the bridge deploys rigid hydraulic outriggers when stationary, and changes to flexible pneumatic tyres when moving.

If the bridge was supported by flexible rubber tyres while heavy trucks were driving over the top of it, it'd probably wobble enough to make everyone involved uncomfortable.

I don’t think ragebol meant that the bridge rolls forward with traffic on it. Just that once a 100m long stretch is finished they can roll the bridge 100m forward with the traffic re-routed or suspended during the repositiong. If they time it right the resurfacing can be done with minimal disruption in the dead of night.

Half right. At night, they direct all traffic onto the shoulder / emergency lane and roll the bridge forward 100m with no traffic going over it at the time. By day, the bridge is stationary, traffic goes over it, and work goes on underneath.

Reply on news.ycombinator.com