I recently spoke with Carsten Brinkschulte, co-founder and CEO of Dryad. Below is part of our conversation about Silvanet and how it addresses the growing global concern about wildfires.
Carsten, tell me a little about yourself, Dryad and your product, Silvanet.
I have been in the telecoms industry for 25 years. I have built three startups and exited three times in the sector, in 4G network infrastructure, mobile email, instant messaging services and device management. I founded Dryad in 2020 with five co-founders. Dryad is what you would call an “impact for profit” company. The mission is to be green, not just as a PR exercise. We want a positive environmental impact, but also profits so we can have a bigger impact.
In 2023, we are launching Silvanet to focus on ultra-early detection of wildfires, as they have a devastating environmental impact, particularly on global warming. Every year, wildfires around the world emit between six and eight billion tons of CO2, which represents 20% of global CO2 emissions.
Our mission is to reduce man-made forest fires. Arson, reckless behaviour, accidents and technical failures are responsible for 80% of fires. We want to prevent the loss of biodiversity and CO2 emissions, but also address the economic losses, as fires cause enormous damage. The lower end of the figures is about $150 billion, but that figure can rise to $800 billion a year, depending on how you look at the statistics.
What is your solution?
Silvanet is a comprehensive solution: sensors, network infrastructure and a cloud platform. We have developed a solar-powered gas sensor that we integrate into the forest – it can be hung from a tree. It is like an electronic nose that can smell fire. There is no need to have an open flame – someone can throw a cigarette and, depending on the wind and other parameters, a nearby sensor should be able to detect it within 30 to 60 minutes.
We are running an integrated AI at the edge of the sensor to distinguish between the odors the sensor is exposed to. When the sensor detects a fire, it sends an alert.
The sensors are solar powered. Solar panels are fairly small, but large enough to power the electronics through a supercapacitor to store energy. It’s not as energy dense as a battery, but it doesn’t have the drawback. Lithium ion would be a dumb idea because it can power itself. We didn’t want to carry a lighter into the woods.
Obviously, there is not much direct sunlight under the trees, but the supercapacitors work well at low temperatures and have no limitations regarding recharging cycles. The entire installation is very efficient. We are careful not to use excess energy.
Also, since we are in the middle of a forest, we don’t normally have 4G or any other type of connectivity, so Silvanet works as an IoT mesh network. We use LoRaWan for communications, which is like Wi-Fi but with lower power and longer range – it can communicate over several kilometers. We have added the mesh topology because LoRaWan doesn’t have mesh. As far as we know, no one else has done this.
The mesh allows us to cover large areas without the need for an electrical supply. Sensors communicate from deep within the forest through the mesh to a border gateway. A cloud platform then captures the data, analyses it in depth and sends alerts to firefighters.
What does the implementation look like?
The density of deployment depends on the client. Typically, we deploy patchy deployments with a focus on high-risk, high-value areas. In remote locations, we place fewer sensors, but in areas such as roads, trails, power lines and railway lines, where most fires originate, we place many more.
Humans don’t start fires in the middle of the forest. They start them on trails where people throw away a cigarette, or in a campfire that grows out of control or isn’t properly extinguished. In all other cases, a fire can be caused by lightning, or a tree falling on a power line, or a train sputtering and starting a grass fire that turns into a forest fire and then a wildfire.
In the end, you end up with variable density. You need one sensor per hectare (about three acres) for a fast detection time, and then one sensor for five hectares in total.
Other solutions include optical satellite systems, which look down from space to spot fires with infrared cameras, or cameras on the ground that can see plumes of smoke rising above trees. All of these systems make sense. Satellites are invaluable for seeing where big fires are headed, but they are late in the game when it comes to detection. Cameras are also good because they are closer to the action.
The fastest are certainly electronic sensors, but they cannot be everywhere. So the ideal would be to deploy all three systems. Cameras have the widest view and satellites have the widest image. Sensor systems can be concentrated in high-risk and high-value areas, such as at the interface, where there are people who set fires but are also affected by them.
Do you have an example?
We have a pilot deployment in Lebanon. The deployment was high-density because it is what is called a wildland-urban interface: there are people living in villages, there is some agricultural activity and forests. It is the highest risk and highest value because if there is a fire, there is a good chance that it will spread and become a conflagration, and then there is a catastrophe.
In the pilot, we detected a small fire within about 30 minutes. Initially, the sensor’s AI calculated a 30% chance of a fire from the gas scans. The wind may have changed as the probability decreased, but about 30 minutes later it detected more smoke and “decided” that it was indeed a fire.
How is business going?
We try to keep prices as low as possible: despite being manufactured in Germany, the price of the sensor is less than €100. We have a service fee to operate the cloud, which is charged annually, but is also low-cost.
Last year we sold 20,000 sensors worldwide. We now have 50 installations in Southern Europe (Greece, Spain and Portugal), in the United States (California, Canada, Chile and South Korea). We have one installation in the UK, with the National Trust. We also have three or four forests in Germany, in Brandenburg, which is very prone to fires and is as dry as a powder keg.
This year, we expect to ship over 100,000 sensors. We are ramping up production to be able to handle that volume. We have the right funding in place with venture capital – we just raised another $5.6 million in mid-March to fuel the growth we are seeing.
The vision is to go beyond fire: once a network is installed in the forest, much more can be done. We are starting to work on additional sensors, such as a fuel moisture sensor that can measure fire risk by measuring the moisture of fuel in the ground, a dendron meter that measures tree growth, and a chainsaw detection device to detect illegal logging.
