Hacker News Reader

Comment by scarejunba | Hacker News Reader

I wonder about the human effects of safety engineering. Say you require a 10x margin because you know builders will dilute it to 5x undetectably (say it’s under layers of other material or something). Every time you attempt to compensate for the dilution of the standard in practice, they’ll undo it. But if you stop, they may continue increasing the dilution. There’s a normalization of deviance. And indeed most buildings will still stand because the standard is attempting to handle six nines so everyone involved building will probably ever build something that fails even if they diluted it to five nines.

How do you encourage not doing that? Enforcement is one angle but are there others? Force the buyers to have insurance and then let the insurance company eat the risk assessment?

View on HN

Well one option is to hold designers, developers, builders and inspectors separately liable for the work and inspections they do for their entire lives, not limited to the period of solvency of the company they did the work for.

We’ll see a return to very conservative designs though, with no 100 storey skyscrapers until they find people actually willing to pay for and perform good work.

Until we are actually sending people bankrupt or jailing them, the corruption will worsen. There has to be some actual penalty for people trying to pass off shonky work.

Does this, in practice, mean quality? The aeroplane industry doesn’t do this and they build safe craft.

Aerospace industry has a deeply engrained culture of safety, and improving from failure. Construction industry has a history of building to bare minimum requirements, organized crime involvement, corruption, and passing the buck when anything goes wrong.

I dunno, I'd say there's a very long tradition of safety in construction.

"If a builder builds a house for someone, and does not construct it properly, and the house which he built falls in and kills its owner, then that builder shall be put to death." (Code of Hammurabi)

I look at that the other way.

The need for effective regulation and proper enforcement come with a long tradition in construction business.

FAA certified mechanics and inspectors are held personally liable regardless of who they are working for and what they are working on.

> But if you stop, they may continue increasing the dilution. There’s a normalization of deviance.

For a very good explanation of the mechanism that causes this kind of normalization of deviance, I recommend Richard Cook's outstanding talk "Resilience In Complex Adaptive Systems"[1].

[1] https://www.youtube.com/watch?v=PGLYEDpNu60

Theres a whole field of structural reliability, and it's where they come up with load and resistance factors for design equations, and material testing requirements. I believe the targets are such that the probability of collapse of any structural component over it's lifetime is less than 10%, if it was designed for a utilization of 1.0. In reality, sizes are bumped up so the probability should be even less.

In countries with strong regulatory requirements, you generally don't have to worry about builders 'diluting' because it is inspected by the design engineer and tested by 3rd party testing companies. It is in the engineers best interest to ensure compliance because they are legally liable in that regard. You can actually be suspended from practicing engineering if you do not comply with field review requirements. If our design calls for 35MPa concrete, samples are taken at the job site, tested, and reported back to the contractor and designer. If they come up short, it gets remediated.

The issue is that components dont build systems. Sure, maybe that get concrete that tests to 35Mpa. But was the embedded steel reinforcement coated properly, or will it rust? Were the balconies and doors sealed? Did anyone wash grout down the drains? And then imagine someone finds a problem, like “leaks” in a sibling comment. Do they revisit the strucutral issues that led to it? Or do they patch it up with a new 3rd party tradie and walk away?

The idea being that your point inspection are finite. And the responsibility will be laundered behind a legal or organizational structure. Youre not going to reach anyone “responsible” for systemic issues. And as soon as you do that youll dampen all relevant investment activity. See the Washington State Condominium Act for a great example.

And yet the UK continues to have mini-scandals :

https://www.bbc.com/news/business-46454844

Not sure about UK, but where I'm from this type of single family home construction discussed in the article doesn't normally require a structural engineer. In that case, you really are depending on the reputation of the builder.

Encouraging is easy - put in jail everyone who knew about failures in the process, from the lowest ranking engineer to the top-management, if they were aware. It will put hundreds of people behind the bars at first, but the rest will be more than happy to follow the safety rules instead.

The process would also become a game of whack a mole with safety regulations, as everyone falls over themselves to point out every real or imagined flaw.

This can also be handled by market forces. The people who expect to buy the apartments in a high-rise like this could collectively hire their own inspectors during construction, which would be answerable to them alone. The obvious way to collectivize this is through insurance.

The problem stated by the linked Reddit commentator happens because the interests of everyone (architects, engineers, constructors and inspectors) currently align to de-prioritize safety.

There's numerous reasons why that doesn't happen. One being that no individual apartments might have been sold before the construction is finished, but another reason could be that building codes have been good enough, so people think of that as a waste of money.

Privately hired inspectors is a terrible idea. That leads it to only the rich being able to afford to buy good inspectors leading to only the rich getting safe housing.

Your whole comment can be condensed down to "I think poor people should have to live in dangerously built housing and die but it'll be their own fault for being poor."

There needs to be much more regulation with much higher penalties for failures, and a dismantlement of the incentives of corruption.

> Your whole comment can be condensed down to "I think poor people should have to live in dangerously built housing and die but it'll be their own fault for being poor."

I think this is accidentally being uncharitable. It's possible that GP did not think all the way through to the conclusion you saw. A more accurate inference that doesn't assume hatred of the poor from GP would be something like "this will lead to a system where only the rich can verify the quality of housing" and stopping there.

Consumers feel like they need to consider the safety rating of cars they buy, and they pay to e.g. buy brands like Volvo or Tesla known for exceeding the bare minimum when it comes to safety.

Yet I don't think you'd say that someone advocating that we continue that system thinks that poor people should drive dangerously built cars, and it'll be their own fault of they die in car accidents.

I'm pointing out that it's interesting that there seems to be no analogous mechanism to ensure safety when it comes to buying houses, which are a much bigger investment than a car.

> There needs to be much more regulation[...]

If the Reddit OP is to be believed the problem isn't that there aren't regulations, but that they're systematically not being enforced.

> [...]with much higher penalties for failures[...]

I doubt that would work as well as aligning incentives. Even if you gave engineers the death penalty for cutting corners they stand to make a lot of money, and most of the time they'd get away with it. How many buildings in Sydney aren't in the news?

> [...]a dismantlement of the incentives of corruption[...]

The best way to dismantle such incentives is to ensure that there's conflicting interests. Here buyers are being screwed over by the builders, certifying authorities and the councils because the interests of all three align to screw buyers over.

> I doubt that would work as well as aligning incentives. Even if you gave engineers the death penalty for cutting corners they stand to make a lot of money, and most of the time they'd get away with it. How many buildings in Sydney aren't in the news?

In the US, Canada, and I presume Australia, the design engineer is actually separate from the contractor. Generally, there is a developer who pays all the design consultants and the contractor to make the project reality. So the engineer really doesn't stand to make money from cutting corners. In fact, they have much to lose, including their ability to continue practicing engineering (it's a regulated profession) Their fee is the same if it gets built correctly or not, and it costs them nothing extra to ensure it's built correctly and mistakes are rectified.

But as a neighbour, I don’t want the building to fall on me. Part of the risk is externalized. Do I get to send an inspector too? That over privileges me because then I can stall projects by claiming an ostensible high threshold of risk.

You couldn't stall a project by holding it to a higher standard than the building codes supposedly being violated (if the Reddit OP is to be believed).

Fair point. If the standards were inflated we'd have to bring them back down (which may be politically infeasible) but what you say sounds fine.

A friend's father was working as a private safety inspector after retiring as a civil engineer.

One day someone came to him, offered him additional payment to ignore cost skimming, and told him he's got a nice family and that it would be a shame if something bad happened to them. He quit the same day.

When you have organized crime involvement market forces cannot be relied upon.

This is why there are phased inspection requirements along the buildout - so, when struct finishes/is in progress, an inspection is required to be passed prior to next phase.

The real problem is in the purchasing of Chinese Steel, which supposedly passes muster, but clearly cannot. Hence, Bay Bridge.

But - IMO - this is what was so fucking shady about taking all the steel from the towers on 9/11 and shipping/selling them off to china.

Then we pull a PR Bullshit move of "incorporating" some of the steel in a fucking destroyer ship as some limp "Freedom Flex"

The incorporation of tower steel into a war vessel was a fuck you from the people who profited off the event.

As far as I remember, many of the steel failures on the Bay Bridge were US steel from Ohio through Dyson.

>The eastern span of the Bay Bridge was completed in 2013, six years behind schedule and billions of dollars over budget. Critics say pressure to contain costs and meet deadlines led to poor management decisions, including the use of substandard ---- Chinese steel. ---- Other problems include cracks in the foundation and missing sections of deck drainage that allow rainwater to bleed inside the structure.

https://www.courthousenews.com/engineer-unfazed-reports-subp...

---

There are other articles about it...

> Less than six months after the ill-fated batch of rods arrived in California in 2008, six high-strength rods that Dyson Corp. of Ohio shipped to rebuild the Hood Canal pontoon bridge — which connects the Kitsap and Olympic peninsulas in the western part of the state — failed within days of being installed, according to a review recently made public by California bridge officials.

Right but there are also articles about Dyson Steel:

https://m.sfgate.com/bayarea/article/Bay-Bridge-supplier-fai...

I chose my words carefully. I didn't intend to say that Chinese Steel is magic. Just that they're not the only culprit. If we'd picked up more Ohio steel we'd probably be in the same situation.

Reply on news.ycombinator.com