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.
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.
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.
Smokey would have a different tone too. "Only you, plus our array of guided missiles, can prevent forest fires."
as i said in original post - "not small quadcopters, more like WWII size bombers". I use V1 as a close and real existing example of a drone wrt. size and simplicity/cheapness of construction. Though i think 2x-4x of that size would be more suitable and convenient.
>how to make the vehicle recoverable/reusable
just regular software for a plane type drone. The drone comes, drops the water and returns. Can do it GPS/inertial, AI targeted or remote controlled (like say put a Starlink on it how Ukrainians do it sometimes).
>else how to avoid causing more fire than you're putting out when the vehicle splatters against terrain
while normally the drone would return back, accidental crashes of course unavoidable. You don't have to carry much fuel as the drones can be spread around the state on say a 200km grid. The fuel tank can be made with automatic foam generator or something like this triggered if/when the drone goes down crashing.
the most dangerous word in software engineering
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.
Both of these produce lots of smoke. Neither can be doused by drones.
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.
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.
An IR camera on drone can be fine-grain controlled from the ground when needed - say zoom-in from a 100km on something detected in the sweep like even a camp-fire so you can confirm whether it is a starting wildfire or just a camp-fire - that of course requires hardware like on those pricey military drones, yet it is still a pocket change in this context. You can also much easily reroute the drone. Again, if you watch the drone war footage there is a lot of scenarios where satellite wouldn't work.
Identification only by drone. Then the dousing is done by conventional means, but dispatched locally.
A class 1 fire truck (the big red ones used in sub/urban areas) and class 3 wildland off road trucks carry around 500 gallons of water which is a drop in the bucket. Fire hydrants on the other hand can supply 250-750 gallons per minute.
The best they can do is try to create firebreaks using bulldozers and other off road vehicles while preemptively dousing properties with water to defend them, because they’ll usually have road access. Doesn’t do a damn when Santa Ana winds are causing 90mph gusts that throw embers for hundreds of feet.
They're the real world "ignore friction, air resistance, &c." of kids physics exercises books.
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.
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.
Again, at that point you’re no longer adding value over the status quo. We don’t have a problem detecting large fires near population centres.
People aren't the only (or largest hotter-than-background, and maybe not hottest-human-sized) thing on your property that a privacy-violating IR sensor drone might want to detect.
> Versus a sensor tuned to an actual fire which is substantially hotter and larger.
If you want an early-detection sensor for fires, they'll be substantially hotter but not substantially larger than an adult human.
It's fun/useful to think of things we can do.
Managing wildfire risks mean letting a certain amount burn every year, including deliberately setting fires to create firebreaks and clearing out areas overdue for it. A fleet of drones constantly finding and stopping wildfires in their early stages would do nothing but set the stage for cataclysmic fires down the road.
He did not. It would be worth re-reading his comment. He's pointing out that we do have technology that could help with containment: quickly identify fires, communicate their location, and dispatch some local water carriers. He's also surmising that the cost of keeping these active would be less than the cost of damages, which could very well be true.
Something like a Reaper drone, which he specifically mentions, works fine in the wind, as do the water carriers, that fly at hundreds of miles an hour, that have been actively helping this whole time.
I think this is probably all true, but probably not the future since it would require a competent state government who embraces tech.
Water carriers fly close to the ground and make sharp turns because they need to pick up water and make targeted drops, so they are heavily affected by the wind. The water carriers weren't able to start flying until late this morning/early this afternoon due to the winds being too strong (>75+mph gusts).
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?
With dry conditions and high winds, that is an extremely slow response to an uncontrolled fire...
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.
The problem is that with these winds and humidity small fires become huge fires before anybody can react. Once they are huge fires, detection isn't the problem.
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).