I am working on an interesting school project aimed at detecting fires in a forest, ideally from a distance of around 20 meters, and I am seeking ideas or suggestions for suitable sensors. Unfortunately, I have not yet found any sensor that meets these specific requirements.
I was considering an alternative approach, focusing on detecting IR/UV emissions rather than directly measuring temperature. For this purpose, I thought about using a sensor the KY-026. However, I have been unable to find information regarding the size of a fire that such a sensor could detect at the max distance.
I would appreciate any thoughts or recommendations on this matter. Thank you in advance!
My first school project was a device that could detect a lighted candle in a darkened room from a distance. It used a parabolic mirror and a photoresistor at the focal point. Maybe you could use that type of mirror for your project and experiment with different detectors at the FP. Just remember that whatever is at the focal point can get very hot, enough to start a fire if you point the mirror at the Sun.
This sounds absurdly hard to detect, and if I wanted to use a camera then I think I would be better off detecting smoke and fire!
The idea would be to make a swarm of these "modules" , and that would let me detect if the fire is probably going "out of control"
What distance are we talking about?
That's a great question: I'd define a forest just as a natural place filled with trees e.g. 10 hectares of Terrain filled with trees. I'd rather want to detect fires by wood and foliage mainly, either by detecting the Infrared Radiation from it or Termperature/Heat or to just detect the smoke that falls down on the ground.
I think it's more of a matter of where to place the module optimally, since completely on the ground doesn't sound optimal.
Food for thought... How will these be powered? How will the data be collected? Will the presence of humans to place, recharge, collect data effect the probability of fire? Will the "detectors" become a better path to ground, attracting lightning? The national forest service places (and pays) people in towers to visually scan the horizon to report weather and forest conditions, and if smoke is observed, several reports will triangulate the source of the smoke.
Every forest fire is going "out of control" if left unattended. Point of origin is the more useful thing to know in your arrangement.
Smoke doesn't "fall" to the ground. It's driven by heat, thus it rises.
There will be smoke before visible flame, almost always, due to smoldering from either careless cigarette smoking, abandoned campfires a lightning strike, or even intentional arson. Forests don't just "burst into flames". Plants contain a lot of moisture, so they give off a lot of light grayish smoke from the smoke combined with steam in the incipient fire stage. It is possible for decaying plant matter, through exothermic reaction, to auto ignite, if the temperature reaches the ignition temperature of dry, dead leaves and such; however, this is probably very rare compared to the other reasons forest fires start (most often because of humans, at about 85%).
Deadfall, unless VERY dry, also contains moisture as the ground dew gets trapped under the top layer of the dead plant material.
The best way to detect forest fires is to detect (a) where lightning struck; (b) where people such as hikers and campers are; and (c) using actual humans in spotting towers whose job it is to just look for smoke (most often) or visible flame.
You might accomplish your goal with enough cameras above the canopy such as drone firewatch (10 hectares is doable) or stationary cameras from an elevated position such as an escarpment that overlooks your woodlot.
The you want to do what is done thousands of times in North America. That is set up cameras on a high location and way above the forest canopy and use moveable cameras to watch for smoke plumes. The cameras are all controlled and linked to a central site where one person can watch for smoke plumes from many locations.
I did not test the maximum distance of my parabolic mirror sensor, but it reliably detected a candle from 20 feet away in a darkened room. I mounted the mirror on a salvaged clock motor so the mirror rotated 360 degrees in one minute. It was a small mirror less than the size of a hand. It was originally a Radio Shack solar cigarette lighter, much like this one sold at Walmart.
Given that forest fires can advance at 20 km/hr, neither you nor your equipment should be within a kilometer of a forest fire. The smoke plume can easily be detected from 50 km.