Using drone imagery and AI to rapidly assess damage after hurricanes and floods | Hacker News Reader

Using drone imagery and AI to rapidly assess damage after hurricanes and floods

stories.tamu.edu

48 pointsrbanffy32 comments

great to learn from the headline that this tech only works for disaster response maps, and isn't usable for other types of maps, like mapping out the front lines of a war

A matter of time, people in power will take care of that.

Then a flight plan will be uploaded to the Tet style drones to carry on their duties.

The tech will be useful both for wars as well as for the disaster recovery efforts after your federal funding is cut down for boycotting the wars.

Isn't that a good thing? or at least not bad?

Having a transparent battlefield doesn't necessitate an increase or decrease in casualties.

If you have it and the enemy doesn't you almost certainly will win. If both of you have it, casualties probably go up.

Why would casualties necessarily go up with surveillance? Every argument for precision targeting can be reversed for evasion.

In Ukraine it’s relatively rare for large numbers of troops to be concentrated, because each side knows its opponents would observe the formation and make it a priority target. This makes something like the battle of the Somme unlikely to be repeated.

In call of duty do casualties go up when both sides have UAVs, compared to when both are without?

>In call of duty do casualties go up when both sides have UAVs, compared to when both are without?

Are there any other games updating their play style to recognize the heavy use of drones in war now?

Arma Reforger has very good mods depicting drone combat like flying fpvs and bomber drones. Bohemia interactive simulations also focuses on drones in their newest warsim release

As someone who's in the field, I hate how drones and robotics are now associated with anything related to wars. It just kills the passion, and now whenever you mention you work in it, do it, or are interested in it, you get that suspicious look and even a knock on your door.

Unless you do drones only on your private property, they are inherently a creepy, invasive technology (even though I think they're super cool and like playing with them, too)

Do you think that airplanes, helicopters, and balloons are also inherently a creepy, invasive technology as well? Because from the perspective of capturing imaging data from the air there is really no functional difference between those and UAS...

Reminds me of a project [0] that we spun up in a few days in a hackathon. We finetuned CLIP to be able to work with satellite imagery. Inference was extremely fast.

[0]: https://github.com/arampacha/CLIP-rsicd

I love that we are doing this but I hate that we aren't fixing the root cause of most natural disasters we will be seeing in the coming centuries.

That’s valid, different people are working on that

Agreed. Though, at some point humanity would likely benefit from seeing itself as just humanity and not "us vs them" on this front.

Are they, though? Last time I checked, CO2 is going up without any "flattening of the curve". And whenever mainstream environmentalists annonce X megawatts added to the grid, a tech company announces that they need Y > X megawatts for AI. Plastic production goes up steadily without cessation every year, and there's still immense amounts of deforestation.

No, people "working" on it are intellectually amusing themselves with technology that could be a workable solution if only everyone actually took action and reduced their consumption, which doesn't need that technology in the first place. Pretty much all mainstream solutions are just psychological salves to make us think we are doing something.

Working on it, yeah right. We simply need to make significant reductions in CO2 output and no one is doing that.

Furthermore, technologies like this will make people less likely to do something about the root problem because it ameliorates it.

There is and always will be sizeable work on it. Current work has been termed "greenhushing" by the economist due to the need to be discreet in America due to a pretty intellectually lacking voterbase. Over a billion will likely starve and that is a moral failure that we will all carry for the rest of our lives but 3 people dying in a car crash is better than 4 people dying in a car crash.

You’ve been given free access to this article from The Economist as a gift. You can open the link five times within seven days. After that it will expire.

The remarkable rise of “greenhushing” https://www.economist.com/business/2025/07/29/the-remarkable...

I'm afraid the gift already expired, perhaps due to others accesing the article, but I appreciate the link nonetheless.

The "rise" of greenhushing sounds a lot more like a decline in green pandering.

> We simply need to make significant reductions in CO2 output

Gee, I wonder why no one's thought about that except you.

Maybe because it's actually incredibly difficult to do

A rather sarcastic reply, but I never intended to be original nor surprising. I'm afraid I had to point out the obvious, and for that I apologize as it is almost painful, but the OP statement of "people are working on it" is also equally painful in how obviously ineffectual it is.

It's ineffectual because the task is impossible under current governance. We'll need to hit rock bottom before anything moves, because all the systems we have in place work against any economically hurtful policies.

Well that is true. No argument there. We do need to hit rock bottom, because humans only respond to immediate pain. We are too stupid to collectively work together.

It's not difficult, we know what to do, and we know how to do it. We simply don't want to do it.

It's incredibly difficult in the same sense that it's "incredibly difficult" for a teenager to clean their bedroom. They could do it, and they will if forced to, but they'll drag their feet as much as possible.

That is right. Worldwide quarantine was also an even more difficult task in some ways if you really think about it but the world was much better at that (even if some countries lagged behind) because there was an immediate pain felt, not unlike a parent threatening the immediate revocation of priveleges. With climate change, there is no such immediate threat to the average high-consumption individual.

Ok. The point was that this drone solution by individuals at one university doesn’t detract from other people’s work and it’s a non sequitur to suggest otherwise.

Well, that point itself is invalid. Because ameliorations such as the mapping solution actually makes it less likely that people will do something about the root cause – same as air conditioning makes it less likely that the rich will care about a warming climate.

Like the current increase in wars and their impact on destroying what is left of the planet, after we all started using paper straws?

Is it really most disasters as in more than 50%?

Aren’t earthquakes completely unrelated to climate change?

And hurricanes have happened long before climate change.

And flooding.

No matter what, this is a good technology to have and develop.

>Is it really most disasters as in more than 50%?

This is an absolute minefield of statistics to get into. 90% of all disasters are influenceable by climate change (not an earthquake or volcanic eruption for example). Of that 90%, almost all are currently exacerbated - either in scope, size, or frequency- by climate change.

The IPCC releases a fairly accessible 20 page summary [0], including high-quality citations, and levels-of-confidence. That document gives a good overview of the scale of this problem. It's not reasonable to say that there are 100% more disasters now than there were before anthropogenic climate change. It is reasonable to say that >50% of natural disasters are noticeably exacerbated by anthropogenic climate change.

[0] https://www.ipcc.ch/report/ar6/wg1/downloads/report/IPCC_AR6...

Also can detect unapproved changes to your home or anything an insurance company believes makes your property more risky to insure.

I always thought drones could be very veeery useful for fires as well (not only floods)

Reply on news.ycombinator.com