5 ms·

One can imagine an IR monitoring (satellite or high flying drones like Reaper, one drone can see an IR source like a tank from 100km, so it would take just a fe

by trhway 2y ago

One can imagine an IR monitoring (satellite or high flying drones like Reaper, one drone can see an IR source like a tank from 100km, so it would take just a few drones to monitor the the whole state for fires) with [almost] automated immediate dispatching of the fleet of drones (not small quadcopters, more like WWII size bombers) once the fire is detected. Would be much cheaper than having multi-billion dollar fires every year (these LA fires have already hit $50B as of today).

jkaptur 2y ago
https://wifire.ucsd.edu/firis-in-depth https://wifire.ucsd.edu/firis-in-depth
- trhway 2y ago
  they use manned planes for monitoring - very costly and not scalable. And there is no "water-bombing" drone fleet. The manned "water-bombers" are extremely expensive, there is only small number of them and they actually carry pretty small amount of water as they are mostly helicopters or retrofitted passenger planes.
  - jkaptur 2y ago
    Knowing how things actually work in reality can help guide the imagination.
  - kristjansson 2y ago
    The problem is not that the water bombers are retrofitted, the problem is that water is heavy, so you're up against the physics of flight. An aircraft loaded to its max weight with water (or water-like material like retardant) is going to look empty, visually. Even e.g. a cargo 747 has ~750 m3 (=~200k gal h2o) of cargo space, but 112000 kg (=~30k gal h2o) of payload, so one could only fill ~1/6 of the cargo volume with water. A drone is up against the same physical constraints - water is heavy, and you need a lot of it to have a material impact on a fire, even one you've caught early.
    - trhway 2y ago
      >Even e.g. a cargo 747 has ~750 m3 (=~200k gal h2o) of cargo space, but 112000 kg (=~30k gal h2o) of payload, so one could only fill ~1/6 of the cargo volume with water. yep. Yet they still have to push through the air and have the mass and the cost of the plane for that unnecessary 6x volume. That is because the plane is retrofitted instead of having been built for the purpose of carrying high density payload. (also that 6x larger than needed volume makes those retrofitted planes much more susceptible to the high wind gusts, etc. which almost always an issue with those fires) CA has C-130 water tankers with 20 ton payload (16 ton is the water itself). Total capacity of C-130 when used as fuel tanker - ie. when its own fuel and payload combined - is 30 ton. Empty it is 35 ton, max - 70 tons, and cost $60M+. Taking a real cheap large drone as an example to have real base numbers to work with - German V1 (pulsejet doesn't matter, the piston engine of course would be the way to go here as we don't need jet speed) - 2.2 ton total weight with 0.5 ton fuel and 0.85 ton payload. The ratio of fuel plus payload to weight is already better than C-130's (at that ratio the C-130 would have to carry 35 ton payload instead of its current 20 and 20 ton fuel instead of its current 10), and with piston engine of the same weight as the pulsejet we'd get even larger thrust - i.e can get even higher payload ratio - while using much less fuel (the pulsejet had atrocious efficiency). So, it would take 15 drones of that size for one C-130. These drones are much simpler and easier to build and thus cheaper than even small Cesna, thus 15 of them would be at least ~10x cheaper than one C-130. Dropping 15 loads of 1+ ton water instead of one 16 ton load seems to be better in most cases as the fire line isn't a straight line, and the large plane like C-130 has to maneuver, etc. while flying low over the fire and being subjected to the fire draft, wind gusts, etc. The 15 drones can drop those 15 loads of 1 ton using dive-bomber style - impossible for C-130 or any other air tanker currently in use - thus avoiding that prolonged horizontal low-flying over fire and thus having much less risk/danger form the wind gusts, etc.
      - kristjansson 2y ago
        I think I (and maybe others) understood 'drone' to mean something closer to quad-copters. Sounds like you're actually proposing something closer to water-filled cruise missiles? Which actually seems like a compelling idea. I'd guess there are some questions around how to make the vehicle recoverable/reusable, or else how to avoid causing more fire than you're putting out when the vehicle splatters against terrain, and even more questions around operating the things (autonomously vs. remotely, launch/recovery, ...). But I do love the image of a satellite spotting a quarter-acre IR hotspot and firing a dozen USFS Tomahawks filled with with retardant to nip it in the bud.
JumpCrisscross 2y ago
> it would take just a few drones per the whole state What are you basing this on? Being able to see something once you know it’s there is a different problem from detecting it in the first place.
- trhway 2y ago
  there is no issue in detecting. You sweep the area. The IR sources like a tank - pops up like Xmas tree even on relatively cheap IR from 10km (there is a lot of footage from Ukraine war for example). A house fire would easily pop-up even from tens of km. A acre size fire - it would pop-up even on coarse grain IR from space.
  - JumpCrisscross 2y ago
    > house fire would easily pop-up even from tens of km. A acre size fire - it would pop-up even on coarse grain IR from space Both of these produce lots of smoke. Neither can be doused by drones.
    - trhway 2y ago
      >Both of these produce lots of smoke. i suggest you view a bunch of real IR footage. > Neither can be doused by drones. I'm pretty sure 20 tons of water would douse a house fire. It would take 10 flights of an F8F Bearcat sized drone. Though the F8F is an overkill built for speed. One can carry 1-2 ton with much simpler plane today for that purpose. An acre size fire - an acre-inch of water is 100 ton. So 10 drones 10 flights at 1 ton/drone.
      - JumpCrisscross 2y ago
        > suggest you view a bunch of real IR footage For what? I’m familiar with very sensitive IR systems. I’m struggling to see the advantage of having them on drones versus satellites, particularly if the threshold is an entire house is on fire.
esalman 2y ago
There's no air support to contain fire because of dangerously high wind. Asking for drones now is like a dumb person's idea of cheap science fiction.
- JumpCrisscross 2y ago
  I think they’re talking about dousing when it first lights. That isn’t “a dumb person’s idea.” But it is presently beyond our capabilities, to say nothing of the huge privacy problem it would entail.
  - Jtsummers 2y ago
    They're talking about both detecting and dousing. Note they mention two different kinds of drones (a few monitor drones, and then a fleet of drones to deploy and douse). The monitor drones could be practical, we've demonstrated that loitering sensor drones work well in other domains. Depending on the sensor payload (what it's looking for) and how the data is stored and used, it wouldn't really entail a huge privacy problem unless you show up on IR sensors as well as a fire.
    - JumpCrisscross 2y ago
      > it wouldn't really entail a huge privacy problem unless you show up on IR sensors as well as a fire If you’re hoping to do better than sighting a smoke column, you will absolutely have to be. We’re not at a loss of detecting entire houses burning.
      - Jtsummers 2y ago
        To violate your privacy an IR sensor would have to be able to pick out people (98 degrees Fahrenheit, approx.) and put them on a map. Versus a sensor tuned to an actual fire which is substantially hotter and larger.
        JumpCrisscross 2y ago
crazygringo 2y ago
You're currently being downvoted, but I'm genuinely curious -- is continuous IR monitoring from the sky an effective way to detect wildfires earlier? Or are small fires indistinguishable from a million other hot objects, until the point where they get large enough to detect reliably, it's already obvious to everyone on the ground from the smoke?
- 0xffff2 2y ago
  Yes, it is effective. We have several NASA/NOAA satellites with IR instruments on them that are used for fire detection. These are obviously much more useful for detecting fires that start in very rural areas under moderate conditions. The current crop of fires in LA are all very close to populated areas and spread fast, so satellite mapping wasn't of any use for detection. Also, California in particular has a really great fleet of IR mapping planes that give firefighters and the public much more detailed fire information than the satellite coverage, but when I lived in Oregon where they don't have a similar system, the satellite coverage was often the most up to date information on fire perimeters.
  - maxerickson 2y ago
    Seems like it would be logistically difficult to be able to drop a significant amount of water anywhere in, say, 15 minutes. With dry conditions and high winds, that is an extremely slow response to an uncontrolled fire...
- atonse 2y ago
  I'm not so sure. In nVidia's keynote, they had an example of small smoke plumes identified as potential fires. If you had a drone hovering over a particular city to monitor fires, it could potentially be an early warning system, no? And to others that said that the winds are dangerous, do quadcopter drones (which tend to be more stable anyway) have algorithms to account for that?
  - trhway 2y ago
    >And to others that said that the winds are dangerous, do quadcopter drones (which tend to be more stable anyway) have algorithms to account for that? while they have such algorithms, a quadcopter wouldn't able to achieve needed stability in 100mph wind. A heavy plane would do much better. The current issue is that all those existing water-carrying planes have to go low over fire - thus wind is a problem - as they can't dive-bomb and can't be retrofitted for that. It is just different types of planes. A dive-bomber would do much better - would deliver the water precisely from a bit higher altitude and from a much more stable, even in the strong wind, trajectory. An unmannend dive-bomber is even easier, can go for even better trajectories.
carabiner 2y ago
Your cost analysis is both thorough and thought-provoking.
gamblor956 2y ago
The problem is not identifying fires. That's the easy part. As I watched the local news last night, I was able to watch (along with the anchors) the real-time growth of the Eaton fire from a tiny fire near the campgrounds into a behemoth. This morning, the local news covered the real-time development of 2 new fires (both were quickly contained). The problem is fighting fires in steep, mountainous terrain filled with dryed out brush and trees when the winds are so strong (hurricane-strength gusts) that you can't provide any air support. The problem is that the winds were so strong last night and this morning that burning embers in the air could fly to and light structures miles away (which is how most of the current fires in Altadena started).

by trhway 2y ago

jkaptur 2y ago

https://wifire.ucsd.edu/firis-in-depth https://wifire.ucsd.edu/firis-in-depth

trhway 2y ago
they use manned planes for monitoring - very costly and not scalable. And there is no "water-bombing" drone fleet. The manned "water-bombers" are extremely expensive, there is only small number of them and they actually carry pretty small amount of water as they are mostly helicopters or retrofitted passenger planes.
- jkaptur 2y ago
  Knowing how things actually work in reality can help guide the imagination.
- kristjansson 2y ago
  The problem is not that the water bombers are retrofitted, the problem is that water is heavy, so you're up against the physics of flight. An aircraft loaded to its max weight with water (or water-like material like retardant) is going to look empty, visually. Even e.g. a cargo 747 has ~750 m3 (=~200k gal h2o) of cargo space, but 112000 kg (=~30k gal h2o) of payload, so one could only fill ~1/6 of the cargo volume with water. A drone is up against the same physical constraints - water is heavy, and you need a lot of it to have a material impact on a fire, even one you've caught early.
  - trhway 2y ago
    >Even e.g. a cargo 747 has ~750 m3 (=~200k gal h2o) of cargo space, but 112000 kg (=~30k gal h2o) of payload, so one could only fill ~1/6 of the cargo volume with water. yep. Yet they still have to push through the air and have the mass and the cost of the plane for that unnecessary 6x volume. That is because the plane is retrofitted instead of having been built for the purpose of carrying high density payload. (also that 6x larger than needed volume makes those retrofitted planes much more susceptible to the high wind gusts, etc. which almost always an issue with those fires) CA has C-130 water tankers with 20 ton payload (16 ton is the water itself). Total capacity of C-130 when used as fuel tanker - ie. when its own fuel and payload combined - is 30 ton. Empty it is 35 ton, max - 70 tons, and cost $60M+. Taking a real cheap large drone as an example to have real base numbers to work with - German V1 (pulsejet doesn't matter, the piston engine of course would be the way to go here as we don't need jet speed) - 2.2 ton total weight with 0.5 ton fuel and 0.85 ton payload. The ratio of fuel plus payload to weight is already better than C-130's (at that ratio the C-130 would have to carry 35 ton payload instead of its current 20 and 20 ton fuel instead of its current 10), and with piston engine of the same weight as the pulsejet we'd get even larger thrust - i.e can get even higher payload ratio - while using much less fuel (the pulsejet had atrocious efficiency). So, it would take 15 drones of that size for one C-130. These drones are much simpler and easier to build and thus cheaper than even small Cesna, thus 15 of them would be at least ~10x cheaper than one C-130. Dropping 15 loads of 1+ ton water instead of one 16 ton load seems to be better in most cases as the fire line isn't a straight line, and the large plane like C-130 has to maneuver, etc. while flying low over the fire and being subjected to the fire draft, wind gusts, etc. The 15 drones can drop those 15 loads of 1 ton using dive-bomber style - impossible for C-130 or any other air tanker currently in use - thus avoiding that prolonged horizontal low-flying over fire and thus having much less risk/danger form the wind gusts, etc.
    - kristjansson 2y ago
      I think I (and maybe others) understood 'drone' to mean something closer to quad-copters. Sounds like you're actually proposing something closer to water-filled cruise missiles? Which actually seems like a compelling idea. I'd guess there are some questions around how to make the vehicle recoverable/reusable, or else how to avoid causing more fire than you're putting out when the vehicle splatters against terrain, and even more questions around operating the things (autonomously vs. remotely, launch/recovery, ...). But I do love the image of a satellite spotting a quarter-acre IR hotspot and firing a dozen USFS Tomahawks filled with with retardant to nip it in the bud.

JumpCrisscross 2y ago

> it would take just a few drones per the whole state What are you basing this on? Being able to see something once you know it’s there is a different problem from detecting it in the first place.

trhway 2y ago
there is no issue in detecting. You sweep the area. The IR sources like a tank - pops up like Xmas tree even on relatively cheap IR from 10km (there is a lot of footage from Ukraine war for example). A house fire would easily pop-up even from tens of km. A acre size fire - it would pop-up even on coarse grain IR from space.
- JumpCrisscross 2y ago
  > house fire would easily pop-up even from tens of km. A acre size fire - it would pop-up even on coarse grain IR from space Both of these produce lots of smoke. Neither can be doused by drones.
  - trhway 2y ago
    >Both of these produce lots of smoke. i suggest you view a bunch of real IR footage. > Neither can be doused by drones. I'm pretty sure 20 tons of water would douse a house fire. It would take 10 flights of an F8F Bearcat sized drone. Though the F8F is an overkill built for speed. One can carry 1-2 ton with much simpler plane today for that purpose. An acre size fire - an acre-inch of water is 100 ton. So 10 drones 10 flights at 1 ton/drone.
    - JumpCrisscross 2y ago
      > suggest you view a bunch of real IR footage For what? I’m familiar with very sensitive IR systems. I’m struggling to see the advantage of having them on drones versus satellites, particularly if the threshold is an entire house is on fire.

esalman 2y ago

There's no air support to contain fire because of dangerously high wind. Asking for drones now is like a dumb person's idea of cheap science fiction.

JumpCrisscross 2y ago
I think they’re talking about dousing when it first lights. That isn’t “a dumb person’s idea.” But it is presently beyond our capabilities, to say nothing of the huge privacy problem it would entail.
- Jtsummers 2y ago
  They're talking about both detecting and dousing. Note they mention two different kinds of drones (a few monitor drones, and then a fleet of drones to deploy and douse). The monitor drones could be practical, we've demonstrated that loitering sensor drones work well in other domains. Depending on the sensor payload (what it's looking for) and how the data is stored and used, it wouldn't really entail a huge privacy problem unless you show up on IR sensors as well as a fire.
  - JumpCrisscross 2y ago
    > it wouldn't really entail a huge privacy problem unless you show up on IR sensors as well as a fire If you’re hoping to do better than sighting a smoke column, you will absolutely have to be. We’re not at a loss of detecting entire houses burning.
    - Jtsummers 2y ago
      To violate your privacy an IR sensor would have to be able to pick out people (98 degrees Fahrenheit, approx.) and put them on a map. Versus a sensor tuned to an actual fire which is substantially hotter and larger.
      - JumpCrisscross 2y ago

crazygringo 2y ago

You're currently being downvoted, but I'm genuinely curious -- is continuous IR monitoring from the sky an effective way to detect wildfires earlier? Or are small fires indistinguishable from a million other hot objects, until the point where they get large enough to detect reliably, it's already obvious to everyone on the ground from the smoke?

0xffff2 2y ago
Yes, it is effective. We have several NASA/NOAA satellites with IR instruments on them that are used for fire detection. These are obviously much more useful for detecting fires that start in very rural areas under moderate conditions. The current crop of fires in LA are all very close to populated areas and spread fast, so satellite mapping wasn't of any use for detection. Also, California in particular has a really great fleet of IR mapping planes that give firefighters and the public much more detailed fire information than the satellite coverage, but when I lived in Oregon where they don't have a similar system, the satellite coverage was often the most up to date information on fire perimeters.
- maxerickson 2y ago
  Seems like it would be logistically difficult to be able to drop a significant amount of water anywhere in, say, 15 minutes. With dry conditions and high winds, that is an extremely slow response to an uncontrolled fire...
atonse 2y ago
I'm not so sure. In nVidia's keynote, they had an example of small smoke plumes identified as potential fires. If you had a drone hovering over a particular city to monitor fires, it could potentially be an early warning system, no? And to others that said that the winds are dangerous, do quadcopter drones (which tend to be more stable anyway) have algorithms to account for that?
- trhway 2y ago
  >And to others that said that the winds are dangerous, do quadcopter drones (which tend to be more stable anyway) have algorithms to account for that? while they have such algorithms, a quadcopter wouldn't able to achieve needed stability in 100mph wind. A heavy plane would do much better. The current issue is that all those existing water-carrying planes have to go low over fire - thus wind is a problem - as they can't dive-bomb and can't be retrofitted for that. It is just different types of planes. A dive-bomber would do much better - would deliver the water precisely from a bit higher altitude and from a much more stable, even in the strong wind, trajectory. An unmannend dive-bomber is even easier, can go for even better trajectories.

carabiner 2y ago

Your cost analysis is both thorough and thought-provoking.

gamblor956 2y ago

The problem is not identifying fires. That's the easy part. As I watched the local news last night, I was able to watch (along with the anchors) the real-time growth of the Eaton fire from a tiny fire near the campgrounds into a behemoth. This morning, the local news covered the real-time development of 2 new fires (both were quickly contained). The problem is fighting fires in steep, mountainous terrain filled with dryed out brush and trees when the winds are so strong (hurricane-strength gusts) that you can't provide any air support. The problem is that the winds were so strong last night and this morning that burning embers in the air could fly to and light structures miles away (which is how most of the current fires in Altadena started).