Fire-Fighting Frontlines: Moonshot Material
The Rise of Drone Technology, Vortex Cannons, and Fire Weather
By Valkyrie Holmes
Note: This is Part 2 of a three-part series surrounding Fighting Fire with Sound. You can check out the first part here!
Firefighters have a hard job. That much is true. But why is it so hard to put out a fire? Why do fires get as big as they do? How could this be improved? All this and more will be answered in this deep dive into the wild world of autonomous drone technology and how that could be used to solve one of the biggest problems in fire-fighting: communication.
The Ins and Outs of Fire
For those of you who are unfamiliar with how fire works, lemme give you a quick rundown. Fire can be described as the result of combustion or burning, in which substances combine chemically with oxygen from the air and typically give out bright light, heat, and smoke. You can think of it as the rapid oxidation of a material that emits various reactive properties.
Before a fire even starts, there have to be three things: oxygen, heat, and fuel. It needs heat for ignition to occur, which depends on the materials since every object has a different flashpoint (the lowest temperature at which they ignite). Fuel is a material that serves as a catalyst for the actual burning aspect. It could be anything from paper to oil to wood to fabrics and is the most difficult side of the triangle to actually remove. Finally, we have oxygen, the sustaining factor that reacts with the burning fuel to release heat and CO2.
Many think that the visible portion of fire, the flame, is the only part when in fact, the actual heat comes from within. Fire is hot to the touch because of the conversion of the weaker bond of oxygen to the stronger bonds of carbon dioxide. The fire releases energy and if hot enough, the gases may become ionized to produce plasma (this is important for later). This is when a substance reaches the minimum amount of energy required to remove an electron from a neutral gaseous atom, in this case, in our oxygen molecule, but the matter would have to be at an extremely high temperature in order for this to happen.
The problem with so much fire is that the conversion of oxygen to carbon dioxide releases CO2 into the air along with other greenhouse gases, which continue to warm the planet. Wildfires are extremely dangerous to us in the fact that they remove HUGE amounts of carbon dioxide from our air, approximately 8 billion metric tons of CO2 per year for the last twenty years! It’s estimated that wildfires make up 5–10% of the annual global emissions. According to NOAA Scientist Pieter Tans,
“a very large, hot fire destroying 500,000 acres would produce as much CO2 as 6 coal power plants each year.”
Fires also create their own weather in the wild; during a wildfire event, the thermals coming off of the fire change the wind, the velocity, the surrounding air, and virtually everything else associated with the environment surrounding it. You get a strong wind that comes off of big fires and in order to compensate for that wind, you need an equal and opposite pressure force to equalize and then an increase in pressure to actually put that fire out. This is super difficult since these fires are creating 60–70mph winds (hurricane intensity) from the front.
Vortex Rings to Manipulate Fires
A vortex ring is generated any time a fluid is forced through a hole quickly and forcefully. This causes the fluid to emerge as a fast-moving “ball of fluid”, which curls back in on itself as it moves through its surrounding medium and forms a ring.
The first viable idea using vortex rings to fight fires was presented in 2008 when college students Seth Robertson and Viet Tran of George Mason University added new technology to the already unsuccessful version patented by the US Defense Advanced Research Projects Industry. They eventually patented the “vortex cannon”, a device that utilizes the same principles along with a speaker to spread air particles across a large area.
The design consists of a few different components. The first is a subwoofer, which is essentially a large air pump for low-frequency oscillations. You can think of it as a speaker. A collimator guides the waves/pulses produced by the subwoofer so that they are parallel as opposed to radial. When fluid (which in this case is just air) flows in a parallel fashion it is said to have a laminar flow. If the fluid flow mixes and follows an unpredictable path, it is said to have turbulent flow.
The two students hooked a red tube and a speaker to the back of a wave guide (collimator) and in the front, they had something called a baffle, which is a smaller opening that the air is forced through. It reduced the diameter of the wave guide from 10 inches to 4 inches so that when the subwoofer pumped air through, it increased pressure within the device. The velocity of the air coming out goes up and the pressure goes down, forcing it to fold on itself.
Viet and Seth went on to partner with ARSAC Technologies to begin creating an integrated system designed to specifically fight large wildfires that rely on arrays of acoustic extinguishers, sensing technologies, and an army of ISR (intelligence, surveillance, and reconnaissance) drones. Their goal has been to stop fires from spreading and aims to not only detect embers but also track the location and direction of burgeoning fires to prevent them from crossing property lines.
The Relationship Between Vortex Rings and Sound
So what’s all this about sound putting out fires?
Well here’s the thing: sound doesn’t really do anything.
Sound is a mechanical, longitudinal wave meaning that if it’s sustained by the same cycles per second, the structure of the wave does the same thing. It doesn’t matter if it’s on a high-powered motor or a portable speaker, the output is always the same. Sound waves are just a way of explaining the cycles per second in hertz but in actuality, we’re actually talking about 45 motions per second in order to create vortex rings.
The vortex is the real star of the show.
In terms of fluid dynamics, we don’t know that much or have many detailed explanations but here’s a short rundown:
The front side of the ring has positive pressure and the backside has negative pressure. As the ring intersects the laminar boundary of the fire, it’s able to push and pull on its surroundings. This is what puts out the fire. It’s able to push and pull the fire plasma repeatedly off of the fuel bed, breaking down that fire triangle and essentially stripping away that base.
After sufficient time (which depends on the size of the fire) this repeated action extinguishes the fire. You could describe it as some form of turbulence, but that’s a long shot considering it’s only the most common form of fluid dynamics that we can somewhat explain.
Now the flip side of that occurs when there is ambient air standing still. The resting air folds back in upon itself because of the inner string that moves at a higher velocity and that creates a vortex ring. We know that vortex rings work for sure but we’re not 100% sure why they do. We do, however, know that because the exit is a circle, the ring brings the outside air into it, becoming increasingly stronger in a specific direction. These diagrams show a model of the drone, with the vortex cannon pointed at a 45-degree angle to account for the draft created by the rotors.
Now here’s the thing:
If the subwoofer produces too many vortex rings too fast, then preceding rings can start to interfere with the ones before them. This effect depends on the size of the vortex cannon, the conditions caused by the fire, and geographic location (as things like pressure and temperature vary in different geographic regions around the world). The specific frequency that works best for different conditions can be determined using AI and knowledge of the environment where the drones will be used.
Drones produce a lot of downdrafts which can also interfere with the vortex rings. To avoid this, the vortex cannon must be aimed at a very specific angle which can vary depending on external factors. The correct angle for each situation can be determined with AI.
Autonomy and Fire Detection with AI
Another thing we need to understand is airspace communication. There isn’t currently any system connecting planes or jets to drones, especially autonomously and it’s a largely undiscovered concept in fire-fighting. Full autonomy requires a few different things:
- It needs to be able to sense the environment / keep track of its state and location
- Should perceive/understand different data sources and be able to make a plan from those choices
- Act ONLY when it’s safe and avoid risky situations involving personal harm or harm to human beings
The system uses a couple of different things to achieve this. Both camera and infrared sensors are used to map out the area and take in separate information as well as radar, ultrasonic, and sometimes inertial navigation systems. They utilize a wireless safety system with communications from a central node (most likely the admin’s computer) and operate on a secure software foundation like Hypervisor or OTA (over-the-air updates). On these networks, drones can not only communicate with the person in charge but other UAVs as well, allowing a swarm to be controlled to do certain tasks in the wildfire space.
Artificial intelligence uses a series of methods to combine large amounts of data and run it through algorithms that piece it all together. This allows the software to “learn” tasks and how to report information back to a central node. With time, the drone would learn to automatically recognize patterns or features through computer vision, which relies on deep learning to identify pictures or video in real-time. When applied to robots, they can process, analyze, and understand as well as capture all sorts of information about their surroundings.
Possibilities and Barriers
ARSAC has been working on this technology for a number of years and has run into numerous problems with reproducing experiments out in the open. First off, drones produce a lot of downdraft because of the rotors, which interrupts the ejection coming from the vortex ring emitters. For this reason, everything has to be positioned in a specific way, which is where AI comes into play.
The AI portion they’ve been testing has not been developed yet and they’re probably 4–7 years out from truly mastering it with the funding they currently have. They have been experimenting with devices on drones that use smaller, lighter pulse generators that work on the laminar boundary layer of the fire. Suchinder P Dhillon, CEO of ARSAC Technologies, also notes that there needs to be some kind of swarm technology in play to move in concert when attacking the fire. There need to be different angles and control systems taking hits at the fire from all sides and they’ve been looking at augmented reality to make this easier.
You may be thinking, “If the drones have the power to detect and contain fires, why not just extinguish them as well?”
Due to the fact that ARSAC drones currently use petroleum as their main fuel source, these devices are potentially mini bombs. When looking at stopping a wildfire, efficiency is higher than anything else. They’ve looked at fuel cells and batteries and solar energy but none have the needed power density to produce kilowatts on the drone. There is a solid-state battery in the works that could potentially be the solution since it isn’t explosive when exposed to air and has the required power density but that’s still far out from being in circulation.
That’s why drones can only contain the fires with maximum efficiency; you point them at a place that is not currently combusting and prevent the chain from occurring, therefore stopping the fire from creating those strong winds in that specific area.
“I can take this technology and put it against a 20k watt energy fire however, once the fire is putting out that energy and it’s affecting the winds and other things, now the energy going into it has to be increased to counter those effects as well” says Dhillon.
There’s also an easily fixable problem of increasing the hertz or pressure too much. When too high, the second and third rings begin to cannibalize the ones before them (like previously mentioned) and that creates a whole different kind of pressure that doesn’t help anyone. No one knows what the specific numbers are for the maximum amount of pressure but it varies based on certain elevations, temperature, the density of the air, etc.
But once we fix all these technical barriers, we’re fine right?
WRONG, PUT YOUR HANDS UP IN THE AIR WHERE I CAN SEE THEM.
We’ve got the law to deal with.
The law is the biggest barrier in making wildfire-fighting technology. A while back at McCarran airport in Las Vegas, someone put a UAV in the sky at the height of the drone craze and got in the way of an airplane’s flight path. Ever since then, the FAA has put tons of regulations on drones in the US, such as the BVLOS (beyond visual line of sight) regulation barrier. This requires every controller of a drone to be able to see it and with AI-controlled devices, they violate the law since controllers are unable to see the devices.
In a fire situation, those restrictions are NOT lifted. If there is a drone in the air above 200 feet within a wildfire’s proximity, the tankers that bring water to the fire cannot be given the authority to launch. Right now, the law permits drones or airplanes but never both and until you can get an integrated air space, drones for fire fighting just isn’t possible. It will take time to change the regulations but nothing’s impossible!
While there are definitely some issues surrounding this topic, there are endless possibilities for growth and opportunity and with enough time and effort, this solution will see the light and help thousands of people fight fires more efficiently.
This is just the beginning.
If you have any questions, feel free to email me at vholmes113@gmail.com.
Special thanks to Jesse Pound, Soliana Fikru, and Taiho H for helping me with research and being the best team to work with on this incredibly difficult challenge. I’m excited to work with you again soon.