Exploring how Dryad Networks, led by CEO and co-founder Carsten Brinkschulte, uses advanced sensor technology and IoT to pioneer early wildfire detection
The inspiration for Dryad Networks came to Carsten Brinkschulte CEO and co-founder in 2019 during a particularly intense period of wildfires which received widespread coverage in the news. After selling a previous company, he found himself at a crossroads, and so decided to utilise his background in telecommunications and technology to address this global issue.
Wildfires contribute to up to 20% of global CO2 emissions, equal to the emissions from all vehicular traffic annually. By mitigating wildfires, Brinkschulte says Dryad’s aim is to substantially reduce carbon emissions, but we can also safeguard biodiversity and minimise financial losses, which range from £180 billion to £900 billion annually.
Silvanet: Creating the Digital Forest
Silvanet is Dryad’s complete solution for early wildfire detection, along with forest management, growth and health and monitoring. Derived from the Latin word for 'forest network,' it represents Dryad’s foundational technology. “We are essentially constructing a network within forests to host various applications, the primary one being wildfire detection,” Brinkschulte explains. “This network utilises sensors capable of detecting the earliest stages of a fire and then communicates the exact location of the fire via geo-coordinates to firefighters.”
The backbone of this system is a large-scale low power, wide area network, which operates similarly to Wi-Fi but is designed for expansive outdoor environments. It has the enhanced capability to establish a mesh network using robotics, allowing for the creation of extensive IoT networks within forested areas. These devices are solar-powered and mounted on trees, creating a communications bubble approximately one to two kilometres in radius around each gateway.
“Within this network, we deploy sensors that are specifically designed to detect the presence of a fire by sensing the gases emitted,” he elaborates. “This integration of advanced sensor technology with IoT infrastructure significantly enhances our ability to detect forest fires swiftly and accurately.”
IoT
Traditional forest fire detection methods predominantly rely on human observation, either by individuals spotting fires from nearby locations and alerting emergency services, or by personnel using binoculars in lookout towers to watch for smoke. There are also technical methods such as cameras that replicate this visual scanning for smoke plumes above the tree canopy, and satellites that can detect fires from space.
Silvanet represents a new approach by employing gas sensors embedded directly in the forest. “The sensors detect the specific gases produced by organic combustion, offering an advantage in early fire detection,” says Brinkschulte. “Traditional satellite technology, while providing an excellent overview for tracking and predicting fire paths, is less effective for detecting small fires due to its high altitude or infrequent passes. Cameras, although useful over large areas, cannot detect fires that are concealed under the tree canopy and are less effective at night.”
Dryad’s gas sensors, on the other hand, are positioned close to potential ignition points on the forest floor—such as near paths, campsites, and along power lines—where human activity might spark fires. “This proximity allows them to detect nascent fires long before they grow large enough to be visible to cameras or satellites,” he explains. “This early detection capability is crucial for enabling quick response, which is vital for controlling fires while they are still manageable.”
Reliable Data
Ensuring the accuracy and reliability of fire detection via IoT sensors, particularly when using gas sensors, presents challenges, primarily due to the complexity of distinguishing genuine fire signatures from other similar signals. To tackle this issue, Dryad has employed machine learning technology directly within the sensors: “The AI is designed to determine whether the detected gases are indeed indicative of a fire or merely other common but unrelated elements, such as exhaust from a passing diesel truck,” says Brinkschulte.
This use of AI at the sensor level, also known as edge computing, is crucial for Dryad’s system. “Since our network operates on a narrowband technology, it cannot support the transmission of large volumes of data,” he clarifies. “By processing data locally within the sensors, we can maintain high accuracy and significantly reduce the likelihood of false positives. This method ensures that our fire detection system is both reliable and effective in real-world conditions.”
Proximity
The effective range of one of Dryad’s sensors for detecting a fire is approximately 100 metres in radius, roughly the size of a football field. This range, says Brinkschulte, allows for early detection, which is crucial in preventing the escalation of fires. The ability of a sensor to detect a fire within this range depends on several environmental factors such as the volume of combustible material, wind speed and direction, and temperature.
“While our sensors can detect fires beyond this range, the fire would need to grow larger and emit more gases to be detectable at greater distances,” says Brinkschulte. “That is why our focus is primarily on early detection within high-risk and high-value areas, particularly in the wildland-urban interface, where accurate and early detection is most critical.”
This is achieved this with low-cost, easy-to-install sensors that are designed for longevity and do not require batteries, reducing the risk of introducing potential ignition sources into the forest. The sensors are compact, about the size of the palm of a hand. The installation process involves simply drilling a small hole into a tree trunk and securing the sensor with a wooden nail – a method designed to minimise impact on the tree and allow the sensor to remain securely in place for up to ten years.
“We use solar power and supercapacitors for energy,” he tells, explaining that the design philosophy ensures that even though each sensor covers a limited area, they are deployed strategically to maximise coverage and resolution.
Environmental Focus
Brinkschulte takes the environmental impact of deploying IoT devices in natural habitats very seriously. “Our sensors and gateways are designed to be as non-invasive as possible, crucially excluding batteries to eliminate the risk of fire hazards,” he stresses. “Instead, we utilise supercapacitors and solar panels to power these devices, which negates the introduction of potentially hazardous materials into the forest.”
The primary goal of the detection devices is to mitigate wildfires, particularly those that are human-induced and occur at the wildland-urban interface where property and infrastructure are most at risk. “By focusing on early detection and prevention, we aim to reduce both the environmental and economic impacts of wildfires,” says Brinkschulte.
Looking Ahead
Dryad’s vision is to expand the capabilities of IoT in environmental monitoring well beyond the focus on fire detection. Future applications include the development of fuel moisture sensors, which will help us assess and manage fire risks by measuring ground moisture. Additionally, the company has plans to monitor other crucial ecological metrics, such as tree growth, soil moisture levels, and even the water flow within tree stems. These advancements will utilise the existing Silvanet network infrastructure, essentially turning it into a 'nervous system' for the digital forest, broadening its impact on global sustainability.
Brinkschulte shares Dryad’s most ambitious goal, what has calls their 'moonshot': To develop an autonomous system for extinguishing fires. “This project hinges on securing the necessary funding, but if successful, we plan to spend the next three years creating an autonomous drone that can actively extinguish fires detected by our network,” he reveals. “This innovation could significantly enhance our ability to respond to and manage forest fires, marking a significant milestone in the integration of IoT technology with environmental preservation and sustainability efforts.”
===
This article first appeared in the International Fire and Safety Journal and is reproduced here with their permission.