Hacker News Reader

Comment by Reedx | Hacker News Reader

How much room for improvement did you net out at? I wonder if there's a point where it'd be significant enough to make inroads...

The recent rebuild of a section of the Bay Bridge took 11 years and went 2,500% over budget. Whereas the original entire bridge was built in 5 years, ahead of schedule and under budget.

And something I just came across: SF allegedly had 6 Salesforce subscriptions at $1M/YR, and not even using them[1].

Given stuff like that, I worry you're right, it may hardly matter how much better a construction company is.

1. https://twitter.com/michelletandler/status/12734043395669934...

View on HN

Part of the costs is our low tolerance for detouring car traffic. Recently, LA metro made waves by shaving 7 months off of their purple line extension schedule, since the lightened pandemic traffic allowed them to dig an open cut along wilshire and install decking over the future station. Beverly Hills fought that closure from happening since the build began, but the pandemic forced their hand and removed any ground for their argument to stand on.

We could save so much money if we didn't focus so much effort on minimizing construction impact to existing traffic. We would rather take a longer, slower, more painful bleed of the public purse, than cheaply ripping the bandaid off and dealing with a 15 minute longer commute for a few months.

Well that is just offloading costs onto the public as tax is it not? At a $15/hr min wage, adding 15 minutes to a round trip commute for 20'000 people/day M-F costs after 7 months $10.5 million dollars too.

That's not a good comparison. You can actually spend the money saved by finishing construction faster. Nobody gets a check for all the supposed "saved time" when you do that math. You can't pool that money together and build a nice neighborhood park -- because there is no money.

What happens in real life is that people largely don't just stick to their same behavior and suffer like lemmings. They time shift a little bit. They find a different route. It's not magic, it's just a complex system adapting to a change in conditions.

"Dollar value of time saved" calculations are a common distraction peddled by lazy engineers who can't actually make the numbers for their project make sense and instead need to dupe the public into supporting a project they want to do.

I am not fond of this cost-based argument - you hear this one (longer commute times priced at hourly wage) and similar ones.

But no company pays me for my commute time. If I take an extra hour commute I don't get that 15 bucks. And that commute time does not go onto GDP either. It's hard to argue the economy lost the time I was doing nothing productive.

And if we really want to take it to the extreme, if it costs the country 15 bucks each hour people are not paid to work, but could have been doing something useful, Avengers Endgame must have cost the country more than it made in ticket sales. Thanos' final revenge:-)

> But no company pays me for my commute time. If I take an extra hour commute I don't get that 15 bucks. And that commute time does not go onto GDP either. It's hard to argue the economy lost the time I was doing nothing productive.

I suppose that's true if you believe that things without a price are worth nothing.

Truly, it's a mystery why people spend time with their families and friends instead of working another job. Such a waste of time.

>But no company pays me for my commute time. If I take an extra hour commute I don't get that 15 bucks.

Pretty much anyone paid to show up in a windowless van and do some task is paid (or their boss is paid) by billable hour. Anything that decreased the ratio of billable hours to hours worked is going to mean they need to charge you more. Obviously it might amortize out to only be pennies on the dollar but when you start adding a pennies on the dollar sized inefficiency to the economy of an entire region it's gonna hurt.

Imagine if you made the day 5min shorter for 50% of the population. That's what crappy transportation infrastructure does (be it car, rail or otherwise).

>And if we really want to take it to the extreme, if it costs the country 15 bucks each hour people are not paid to work, but could have been doing something useful, Avengers Endgame must have cost the country more than it made in ticket sales. Thanos' final revenge:-)

Is there a word for when you actually agree with a reductio ad absurdem? Because man, what a waste of time movies like that are.

> If I take an extra hour commute I don't get that 15 bucks. And that commute time does not go onto GDP either. It's hard to argue the economy lost the time I was doing nothing productive.

Presumably on average you work an hour's less overtime/day while the diversion is in place, or maybe you quit your job and take a less productive one that's in decent commuting range, or so on.

> And if we really want to take it to the extreme, if it costs the country 15 bucks each hour people are not paid to work, but could have been doing something useful, Avengers Endgame must have cost the country more than it made in ticket sales. Thanos' final revenge:-)

I mean it's certainly worth thinking about how much time you're putting into things like movies and whether that's worth it to you. If something that takes 1 hour and costs $45 is more fun than something that takes 3 hours and costs $15, you might well be better off doing the former.

Spending $10.5 million to save $6 billion sounds like a steal.

Yep, close to 50% of the cost for burying power lines is having to manage and redirect traffic (in Canada)

It's not just a car detour issue. When a city shuts down a street for months to do construction that is likely to destroy any adjacent restaurants and retail businesses. When customers can't easily drive to business and park nearby then they stop coming. Naturally business owners strongly object to this and advocate for slower, less disruptive construction techniques.

This is a valid complaint, but the problem is that slower and less disruptive techniques are still disruptive and still often kill businesses. In many cases it would be much faster and more cost effective to give transfer payments to the businesses for their expected revenues during the closure and then just have them close temporarily to get the project done a lot easier and faster.

No one going to any store in LA is able to park in front of it. Even if there were a street parking spot at that artery, it will probably be occupied by someone not going to that restaurant or store. Plus the sidewalks don't close when the city does this sort of work, and there are ample parking garages in every neighborhood with a commercial corridor that remain open. I saw a figure that said LA had 3x as many parking spaces as cars.

Offer a 10% tax free finders fee to anyone who finds any wasteful spending like that. $600k once for a $6 million annual savings (and that's before you fire whomever is in charge of auditing the books already) is a steal. Accountants and auditors could work like bug bounty hunters.

This is a great idea, but waste is in the eye of the beholder. A $6M environmental impact study might be considered waste by me, considered politically nice by the politician who agreed to it, considered legally necessary by the government's lawyer, and considered vital by the contractor paid to perform the study. Who's right?

If similar environmental impact studies cost half as much or if the study was not read by those who ordered it that is a pretty clear case. Some things are ambiguous I admit but I'm willing to bet there is enough low hanging fruit to keep a good number of people employed looking for it.

Edit: This is a HN faux pas however, please don't downvote the person above, they added to the discussion. I suspect that may be a common objection and I'm glad to be able to clarify it.

If similar environmental impact studies cost half as much...

Who defines "similar"?

Is a study performed in Texas similar to one performed in California? Is a study performed on land that drains into the municipal water supply similar to one performed on land that doesn't? Are two studies performed near each other location 10 years apart "similar" if new laws have come into existence, or a local species has been declared endangered in the meantime?

No matter how obvious the matter might seem, people will still disagree. And the people spending money can always come up with justifications.

...or if the study was not read by those who ordered it that is a pretty clear case.

How do you prove that those who ordered it did not read the study? And what if they did not read the study, but instead read a summary of it prepared by someone qualified who did read the study?

> Who defines "similar"?

A court of law.

> Is a study performed in Texas similar to one performed in California? Is a study performed on land that drains into the municipal water supply similar to one performed on land that doesn't? Are two studies performed near each other location 10 years apart "similar" if new laws have come into existence, or a local species has been declared endangered in the meantime?

I don't have the answer to these questions but I wager that we can find experts who do. Presumably firms exist that do hundreds or thousands of studies annually, those firms could testify how they would bid such a project. What the cost structure typically looks like.

> How do you prove that those who ordered it did not read the study?

A government employee could send a email to the effect that the report was not considered that is later discovered. Alternatively, a whistle-blower may clandestinely gather evidence in order to collect the bounty for themselves.

> And what if they did not read the study, but instead read a summary of it prepared by someone qualified who did read the study?

That seems like a reasonable use of funds.

This isn't going to catch every case, the purpose is to provide a natural check to government corruption and waste. Discoveries made by this system may lead to the judicial branch investigating those who made these choices. The FBI could use these tips to target sting operations. The voting public may use this information to decide whom they are going to elect, and the best part is it's free.

A court of law.

If you think these projects are overpriced now, you can’t imagine how expensive they’ll be when every private citizen has been armed with legislation that lets them take the city to court in an attempt to win these cases. Every project will need a team of lawyers and auditors to check that absolutely everything in the project is beyond reproach. It’s just insane how much things would blow up.

Exactly.

In fact https://www.brookings.edu/wp-content/uploads/2019/07/2019-07... argues that most of the increase in construction costs since the 1960s can be attributed to an increasing role for "citizen voices". I have to believe that increasing the power of citizens to impact the process will only make it worse.

You want to involve a court of law in order to bring costs down? You've just massively increased the project budget and timeline without solving anything.

https://en.m.wikipedia.org/wiki/Cobra_effect

That's an interesting point but it's arguably not the same because you should be able to enforce a strict separation between parties such that it isn't possible for the people who collect bounties to create the waste/corruption that they are being paid the bounty to identify.

In theory, of course. But when there's money involved, it has a funny way of flowing towards "value creation". In theory, you should be able to keep politicians from benefiting from a quid pro quo with lobbyists. In practice, politicians seem to frequently attain cushy jobs in private industry after their terms end (and sometimes before another begins).

Humans are unfailingly creative. I don't know how they'd game the system you set up, but I assure you they will!

Whereas the original entire bridge was built in 5 years, ahead of schedule and under budget.

The original Bay Bridge also had 28 fatalities during its construction. Imagine that happening today!

Even if safety has that magnitude of penalty, it doesn't explain years of delays and 2500% over budget. It would've been factored into the original timeline and budget.

And surely after 80 years of improvements in technology, materials, process, etc we have the ability to build in less time AND have no deaths?

And surely after 80 years of improvements in technology, materials, process, etc we have the ability to build in less time AND have no deaths?

I don’t think this follows. MRI machines save lives. MRI machines did not exist a century ago. MRI machines have made healthcare more expensive, not less.

I think the same could apply to buildings with new materials and construction methods.

Not all healthcare requires MRI machines, no? Presumably most (not all, I will agree) of the marginal expense is shouldered by situations which make use of the MRI, versus those that could carry on without it (plus another small concession for those situations that use the MRI that would’ve resulted in similar outcomes pre-MRI invention).

I agree with your point generally though. It reminds me of the “websites got slower faster than computers got faster, thanks to computers” argument.

I'm not sure that harnesses and hard hats would have delayed that timeline by very much.

I do electronics repair on an offshore oil rig, so not quite the same as construction. When I'm working, using a harness usually doubles to quadruples the time it takes. There is extra paperwork, you need to verify that what you are tying-off to can support the dynamic shock load of a person falling, they tell us 5000 lbfs. We have to put together a plan for if we do fall how the person going to rescue us is going to retrieve us. Also, we can't work alone because if you fall someone needs to be able to notify the rescue team. A 30 second job for one person on the ground may take 2 people an hour when the job is at heights. If you aren't working from scaffolding, you may be working from an unstable work position. There is also intolerance for dropping anything, even if the area under the work area is barricaded off, so each move and fastener removed requires much more concentration. We use special no-drop-tools that are all on lanyards that no matter what sort of organizational strategy frequently end up in a knot.

> There is also intolerance for dropping anything, even if the area under the work area is barricaded off

What’s the reason for this?

It really comes down to everything is a trade-off of some sort. The companies hiring us look at our safety records as part of the vetting process, no one wants their name in the news for bad reasons or dealing with the liability that comes with a worker being injured. Also, nothing slows down a job like an accident investigation.

Some of it is depending on how high up you are something can bounce pretty far, possibly outside the barricaded area. The term used is ALARP- As Low As Reasonably Practicable. The trade off is, when someone's working really high up you can make everyone else stay inside but this wouldn't really be productive so you barricade off directly under where the work is going to take place and where you could anticipate something bouncing too, then do everything to still prevent dropping something. I mentioned with the tools, everything needs to be secured, so for example if I'm removing a CCTV camera, I need to attach some form of attachment to the camera before un-bolting it. Of course it's still not possible to secure the nuts or bolts holding the equipment in place so move slowly, pay close attention to every move made. If something does get dropped, there is more paperwork to do and everyone in the company is notified. They calculate what the potential harm to someone is [1].

[1] https://www.preventdrops.com/safety-regulations/understandin...

I'd bet a dropped tool on an offshore rig could end up causing an expensive breakdown or even explosion. Similar to why aircraft grade stuff needs tool checking and checkouts and all bolts use safety wire, where for a car there is less need as the failure modes are less catastrophic.

It takes more than hard hats and harnesses to ensure safety. Of the 11 fatalities that occurred in the Golden Gate Bridge construction (which took place at the same time as the Bay Bridge), 10 of them occurred simultaneously when a platform collapsed.

In large-scale construction, safety needs to be built into every aspect of planning and execution of the project. That can add large costs.

An expectation that acceptable deaths/injuries = 0 definitely affects the timeline.

We haven't won that battle yet. A worker passed away recently at the sofi stadium construction site, due to a miscommunication on maintaining a panel. I don't think preventing this unfortunate death, which would have just required one person mentioning to another that they are working in a particular area, would have added high costs to the build. I'm curious to know the actual costs of worker safety. It's always painted as the boogyman of the high budget project, but how expensive can PPE, building hasty wooden railings, and importantly communicating danger to employees which was the failure in this case really be? I must be missing some key costs here, and would love to be more informed.

https://www.latimes.com/sports/story/2020-06-12/attorney-for...

Safety is waaaay more about worksite practices and culture than how much you spent on safety equipment.

Source: I worked on a deployed aircraft carrier, one of the most dangerous workplaces in the world.

Any solution to safety problems that relies on people just never making a mistake ever isn't realistically a solution at all. What you need is some process to ensure that when work like this is carried out, the area is appropriately marked as unsafe, and processes add cost. In this case, safety would probably have dictated not carrying out the work on top of the roof whilst the work underneath was in progress due to the risk of them misjudging where the weak area started and falling through, potentially blowing up the schedule or even meaning workers were paid to stand around doing nothing, which adds more cost.

It isn't just worker safety that has improved in the last century, it's the safety of people in the building throughout its life. Building codes have expanded in scope and complexity over the years. Part of their scope includes installation method restricts and worker training and qualifications.

The other aspect of it is general workplace safety regulations which have expanded over the years as well. When budgeting a project, you need to consider both, however. And workers also benefit from both. For example, the restriction on the use of asbestos in building insulation was brought in to reduce the health hazards of living or working in such a building but the construction workers handling the asbestos are clearly the most at-risk group which benefits from the regulation.

>The recent rebuild of a section of the Bay Bridge took 11 years and went 2,500% over budget. Whereas the original entire bridge was built in 5 years, ahead of schedule and under budget.

To use an analogy from heart surgery:

Fixing a car is easy.

Fixing a running car is hard.

There comes a point at which starting from scratch is better than trying to fix the original mess. Which in our business means the company goes out of business and is replaced by something new.

If an earthquake knocked it down somehow people managed to move on until it was rebuilt. The same for the I-85 collapse in Atlanta. Amazingly, traffic was slightly worse but it’s always disastrous so it wasn’t that much worse. Freeway projects should probably shut down the road more often and just get done.

Reply on news.ycombinator.com