Hacker News Board Hacker News, browsed like a classic message board
Live data from Hacker News

AI for American-produced cement and concrete

engineering.fb.com

Post reply on HN
Print view View original

81–90 of 127 posts

Re: AI for American-produced cement and concrete

#81

Post by rogerrogerr »

no_shadowban_6 wrote: post #67
romaniv wrote: post #50

Earlier quoted context omitted.

The current strategy of the AI hype machine is to exhaust people's reserves of attention by presenting a never-ending stream of hard-to-verify "positive" claims. It's Gish Gallop done on the Internet scale with a never-ending parade of tech influencers, proxy "journalists" and low-value accounts. The whole strategy aims for saturation and demoralized acceptance. It's no surprise that people readjust their immediate r…

Written like someone who hasn't used AI since the great paradigm shift of December 2025.

Was that the one immediately after the great paradigm shift of November 2025, and before the great paradigm shift of January 2026? I think I remember it.

Re: AI for American-produced cement and concrete

#82

Post by bluGill »

munk-a wrote: post #69
Animats wrote: post #28

Hand-held devices for testing concrete properties would be more useful. Most concrete problems come from a bad mix - too much water, not enough cement, etc. Concrete testing usually involves cutting a core out of the poured slab and sending it to a lab. Something where you stick a probe in the mix and can reject it before pouring would help. Here are some on-site concrete testers.[1] They're heavy and a pain to use.…

> too much water, not enough cement, etc. I wanted to mention that Concrete is far more complex and regional than folks might imagine. The quality of gravel and sand, local impurities - these all contribute massively. It's probably best to think of it like a wine's terroir - except, unlike a bottle of wine, it's prohibitively expensive to ship both the components and the finalized mixture to different areas. If a reg…

> it's prohibitively expensive to ship both the components and the finalized mixture to different areas.

We could do this if it is important. There are mines in Wisconsin the export sand to the middle east because that is known to work well for fracking and they don't want to risk a local sand not working well. (AFAIK they have never tested local sand properties, but it is possible they have and it doesn't work). In this case the value of the "perfect" is well worth the high shipping costs.

Re: AI for American-produced cement and concrete

#83

Post by givemeethekeys »

sidewndr46 wrote: post #45
MisterTea wrote: post #41

Earlier quoted context omitted.

On-site, before pouring, they use the slump test: https://en.wikipedia.org/wiki/Concrete_slump_test

Glad to see someone pointed this out. The test consists of a bucket, plywood board & a stopwatch.

Seems simple. Is it effective?

Re: AI for American-produced cement and concrete

#84

Post by bluGill »

jauntywundrkind wrote: post #54
sidewndr46 wrote: post #45

Earlier quoted context omitted.

Glad to see someone pointed this out. The test consists of a bucket, plywood board & a stopwatch.

Hopefully there's good empirical data powering the model here, which just added slump prediction: > Alongside the event, Meta is releasing a new AI model for designing concrete mixes, Bayesian Optimization for Concrete (BOxCrete). BOxCrete improves over Meta’s previous models with more robustness to noisy data as well as new features including the ability to predict concrete slump (an important indicator of concrete…

There is a lot of expense in the wasted concrete from all the different pours that are slump tested. There is a lot of cost from concrete that leaves the plant only to fail slump testing when it gets there - not only do you have to empty the truck someplace else, but there maybe contract provisions if you fail to keep the workers busy. Often more than one test is done - if the plant has an order that they know will be tested they test before it leaves the plant (if it fails they can redirect to a different customer who will knowingly accept lesser concrete - but concrete cures on the truck so it would be unacceptable if rejected at the site to go elsewhere). Many smaller jobs skip the test, but they would like it done if it was free.

That said, I'm not sure if the value can ever be greater than a slump test just before pouring.

Re: AI for American-produced cement and concrete

#85

Post by Hasz »

man, I fucking love concrete. So cool. If you like concrete, check out [Tyler Ley's youtube channel](https://www.youtube.com/@TylerLey/videos). Anyways, I am curious why Meta is investing in this. They use a lot of concrete no doubt, but it's nothing compared to a highway, all the commercial builds, etc. I would have expected them to release something like a very general materials optimization framework to explore a space, but glad to see it applied specifically to concrete.

There are a lot of alternative cements to portland, interested to see if that is in-scope. The list of admixtures is also very long and also fairly secretive. UHPC is a pretty cool development, and I am especially bullish about removing rebar and replacing it with FRP bar to limit the eventual rust cracking that comes with the gradual march of carbonation.

Anyways, very cool and looking forward to the mix developments that come out of this framework.

Re: AI for American-produced cement and concrete

#86

Post by bluGill »

harimau777 wrote: post #32
Animats wrote: post #28

Hand-held devices for testing concrete properties would be more useful. Most concrete problems come from a bad mix - too much water, not enough cement, etc. Concrete testing usually involves cutting a core out of the poured slab and sending it to a lab. Something where you stick a probe in the mix and can reject it before pouring would help. Here are some on-site concrete testers.[1] They're heavy and a pain to use.…

I'm surprised the ratios for a given situation isn't standardized by now. Is it just people cutting corners?

My grandpa used to be a concrete inspector (for the state of Minnesota - if you ever drive i394 there he was one of the inspectors for that). Different plants within a normal commute of his house often had very different sand and so needed a different mix.

Re: AI for American-produced cement and concrete

#87

Post by bluGill »

m4rkuskk wrote: post #46
ortusdux wrote: post #15

Tangentially related, but there is a new generation of trucks that mix the concrete on-site. They can output small batches and change the mix on the fly. They solve a lot of headaches! https://cementech.com/volumetric-technology/

This may work on a small scale, not in most commercial use cases. A typical deck pour (400cy) will pour at 70-80cy/hr. you got 9-10cy/truck. Meaning you have 7 to 8 minutes to back in the truck, empty it into the hopper and leave. You barely have time to add water to the mix. Most high-volume concrete plants are "dry-batch", which means all the ingredients get dumped into the drum and the concrete will get mixed whil…

In the large road projects I've seen they bring a concrete plant to the job. Buildings still get trucks coming in.

Re: AI for American-produced cement and concrete

#88

Post by munk-a »

bluGill wrote: post #82
munk-a wrote: post #69

Earlier quoted context omitted.

> too much water, not enough cement, etc. I wanted to mention that Concrete is far more complex and regional than folks might imagine. The quality of gravel and sand, local impurities - these all contribute massively. It's probably best to think of it like a wine's terroir - except, unlike a bottle of wine, it's prohibitively expensive to ship both the components and the finalized mixture to different areas. If a reg…

> it's prohibitively expensive to ship both the components and the finalized mixture to different areas. We could do this if it is important. There are mines in Wisconsin the export sand to the middle east because that is known to work well for fracking and they don't want to risk a local sand not working well. (AFAIK they have never tested local sand properties, but it is possible they have and it doesn't work). In…

We certainly could - it's absolutely possible. The question is if it's economical and so far the market has ruled in most cases that it isn't. Either the project doesn't need such a perfect amalgamation of materials (maybe there is an expected deprecation that doesn't justify such an outlay - possibly earthquake risk would minimize any expected lifespan gain - possibly the materials contractor just can't internally justify the added material cost while remaining attractive to local contractors).

It's all a balance. Imagine a scenario where you can ship in specialized materials to build a bridge with an expected lifespan of 100 years and it'll cost 50M - or you could use local concrete that has an expected lifespan of 15 years and materials would cost 5M. This is a vast simplification of the math but, assuming those expected costs it'd be cheaper to build using local materials and just schedule replacement every 15 years. And, of course, there'll be egg on your face if you build the 50M bridge and then suffer a massive tsunami in two years that destroys the foundations anyways.

To paraphrase a Grady quote: "Engineering isn't a study of building the best thing - it's optimizing the quality we can get for the cost outlay."

Re: AI for American-produced cement and concrete

#89

Post by camillomiller »

sebastianeament wrote: post #23
kevin_thibedeau wrote: post #9

> As a result, producers need a way to rapidly explore and validate new formulations without spending months in the lab. How do you bypass the normal process of pouring test articles and testing them months and years after cure? This is fundamentally a research activity that needs to conduct verifiable science. Not something you can guess at with an LLM.

Hi, I developed the model. We are not bypassing the regular testing process, and are not using LLMs, but Gaussian processes with vetted test data. The predictions are used as recommendations for onsite testing, to accelerate finding mixtures with optimal strength-speed-sustainability trade-offs.

[flagged]

Re: AI for American-produced cement and concrete

#90

Post by prpl »

Animats wrote: post #28

Hand-held devices for testing concrete properties would be more useful. Most concrete problems come from a bad mix - too much water, not enough cement, etc. Concrete testing usually involves cutting a core out of the poured slab and sending it to a lab. Something where you stick a probe in the mix and can reject it before pouring would help. Here are some on-site concrete testers.[1] They're heavy and a pain to use.…

I think people rarely reject even if it fails a slump test though
Post reply on HN