‘Digital Noses’ in the Forest: How Sensors and AI Detect Fires

Sensor von Dryad am Baum.
© Dryad

Start-up Dryad's sensor system sounds the alarm in the event of a forest fire, stopping them before they become difficult to control.

Author Alexandra Rauscher:

Translation Kezia Rice, 10.27.25

In 2024, the concentration of CO2 in the atmosphere rose more dramatically than at any time since measurements began in 1957. This was largely due to continued high emissions from fossil fuels, but the large number of forest fires also played a part. Increased heat and frequent droughts caused by the climate crisis are leading to an increase in fires. This trend continued in 2025; in Europe, forest fires released 12.9 megatonnes of CO2 this year alone. This is also a new record high. A large proportion of the emissions were caused by fires in Spain and Portugal in August, with further fires in Turkey, Cyprus and the Balkans.

Due to harsher environmental conditions, forest fires are prone to spreading rapidly, especially in southern countries. If the source of the fire is deep in the forest, fires are often detected late, which makes them more difficult to fight. Brandenburg-based start-up Dryad has developed a solution to detect forest fires as soon as possible. With the help of sensors and AI, Dryad gives forests their own ‘digital nose’ that sounds the alarm in the event of a fire. Firefighters can then take action, helping to prevent it from spreading.

How does Dryad’s ‘digital nose’ work?

When Dryad CEO Carsten Brinkschulte talks about the ‘digitalisation of the forest’, it can trigger mixed feelings. Digitalisation has now penetrated virtually all areas of life. For many people, the forest has remained a place of retreat, so far largely free of digital technology. However, if you take a closer look at Dryad’s solution, it quickly becomes clear that it involves minimal visible changes to the forest habitat, and at the same time has great potential to protect it.

Dryad’s ‘digital nose’ is a sensor that uses AI to analyse the chemical composition of gases in the air. It detects gases such as hydrogen and carbon monoxide even in very small quantities. The associated AI has been trained to recognise fire indicators from sensor data. This could be, for example, a sudden increase in carbon monoxide in the air. This enables the sensor to ‘smell’ smoke and fire as early as the smouldering phase, even before open flames appear. The AI-based model also makes it possible to assign odours detected by the sensor to different types of burning wood.

Digital infrastructure in the forest

The Dryad sensor is attached to a tree trunk using a wooden nail. A cloud platform connects the sensors to a digital network infrastructure. This requires approximately one sensor per hectare of forest area. There are currently over 20,000 sensors in use worldwide. If a sensor detects a smoke gas pattern, the signal is forwarded via a mesh gateway and a border gateway. This automatically alerts emergency control centres by radio. In addition to gas patterns, the sensor also measures temperature, humidity and air pressure.

Das Sensornetzwerk der digitalen Nasen.
© Dryad

In Lebanon and Thailand, Dryad technology has already been used to detect and combat forest fires in their early stages.

For use in forests, Dryad sensors must be able to withstand harsh weather conditions and animal contact. At the same time, they must consume very little energy. Dryad therefore relies on solar panels and so-called supercapacitors instead of batteries. This also makes the sensors resource-efficient and environmentally friendly. In addition, Dryad says it carefully selects its suppliers and reviews its operating procedures against strict environmental criteria to minimise its own environmental footprint.

Early warning system for fighting forest fires

The advantage of the sensors lies in the time they save. The earlier a fire is detected, the faster firefighters can combat its spread. By being deployed on the ground, in real time and in conjunction with AI, Dryad sensors monitor forests during the critical phase of fire development. This can be decisive in restricting the extent of fire damage. The ‘digital nose’ thus contributes to an effective early warning system for fires in conjunction with other technologies.

Verschiedene Reaktionszeiten von Sensoren, Kameras und Satelliten bei Waldbränden.
© Dryad

From a ‘digital nose’ to a ‘digital ear’

Dryad connects the forest, thus creating the basis for integrating even more data sources, refining analyses and detecting fire risks earlier. Used correctly, networked systems that combine sensors, satellites, imaging and AI can protect forests and contribute to reducing CO2 emissions.

However, the possible applications of Dryad technology are not limited to fighting forest fires. “Once you have a digital network infrastructure in the forest, the possibilities are endless,” emphasises CEO Carsten Brinkschulte. Explaining the start-up’s further plans, he says: “We have a digital nose in the forest—now we’re building a digital ear.” They plan to use microphones to recognise sound patterns to combat poaching and illegal logging, similarly to the project Rainforest Connection. Of course, data protection should be taken into account, especially in the case of sound recognition.

There are also interesting forest protection projects currently underway in Hesse and Bavaria:

In Hesse, researchers are testing how drones and AI can detect forest fires at an early stage. Drones provide real-time images of controlled fires, which are then analysed using AI.

In Bavaria, scientists at the Technical University of Munich have developed a data-based simulation model (‘digital twin’) for the Berchtesgaden National Park, which analyses the long-term consequences of climate change on the forest.

Looking at the Entire Life Cycle: Tips for Sustainable AI Development and Use

AI users and developers can influence AI sustainability. We present the essential steps to creating a sustainable AI model.

Could AI-Powered Robots Be the Answer to Europe’s E-Waste Problem?

A team of researchers have developed AI-powered robots to tackle Europe's gigantic e-waste recycling crisis.

How To Avoid AI Features: Open-Source Software Lets You Escape the ‘Walled Garden’

Carbon-hungry AI is everywhere. We explain how to avoid AI features and why open-source software and decentralised networks can help.

Sustainable AI Means Looking Beyond Data Centres

The massive impact of AI on people and the environment is just the tip of the iceberg. True AI sustainability requires a life-cycle approach focused on the common good.

Die Satellitenaufnahme zeigt die Blaualgenblüte in der Ostsee.
ESA
Algae Monitor: Measuring Buoys and Satellites Help Protect Rivers and Waterways

AI-powered analysis of satellite images and local data helps us detect and counter changes in algae ecosystems early on.

Renewable Energy and Waste Heat Re-Use: Supercomputer MeluXina’s Strategies for Green, Fast Computing Power

Our modern world relies on carbon-emitting supercomputers. Enter: MeluXina, a supercomputer setting the standard for green computing power.

Does Ocean “Dark Matter” Hold the Key to Better Marine Protection?

The AI MareExplore project uses AI to identify marine enzymes that break down plastic and bind CO2.

Are There Sustainable Language Models? Three Perspectives From the Current Debate on AI and GenAI

Will AI save the climate? Or are applications like ChatGPT just CO2 guzzlers? We take a look at three perspectives on language models.