Drones with hypersensitive sensors help researchers map forest health

Annette Eilert

Doctoral student in forestry and wood technology

Annette Eilert took the course Sustainable small-scale forestry at Linnaeus University, followed by a bachelor’s and a master’s degree in forestry and wood. She is now a doctoral student, researching the development of digital solutions to detect damage to pine trees at an early stage using various digital tools.

– The drone’s camera is my eyes in the sky, and the drone’s role is to carry various types of sensors. It’s the sensors that contain the technology, and they’re advancing at lightning speed. They’re like different tools in a toolbox. I choose the tool based on what I want to find out.

Annette Eilert is a doctoral student at the Department of Forestry and Wood Technology at Linnaeus University. Her research focuses on developing methods to detect damage to pine trees at an early stage using drones.

From saplings to mature forests

With sensors mounted on drones, researchers can study trees at all stages, from saplings to mature forests. A certain type of sensor can measure the volume of a forest stand, while another can analyse the distribution of different tree species within the stand, such as pine, spruce, and birch.

Yet other types can measure factors such as moisture levels, emissions of harmful substances, and temperature changes in the trees.

So far in her doctoral project, Annette has primarily used two types of sensors in an indoor laboratory setting: a laser sensor that enables precise measurement of the structure and shape of plants by scanning them with high accuracy, and a multispectral sensor that measures different wavelengths of light reflected from the plant, including visible and near-infrared light. This is used to assess the chlorophyll content and health of the plants.

– I use the sensors to examine the condition of the plants. The sensors allow us to measure what’s called reflectance, meaning the light reflected from the plant’s needles or leaves, which provides us with information about its health.

Growing fleet with great capacity

In 2023, Linnaeus University began building its drone fleet in earnest – a fleet that is now one of the largest for forestry research in Sweden. This initiative has been made possible through The Bridge partnership, a collaboration between Linnaeus University, Södra and Ikea, the DISA frontier research environment in computer science, and external research funders.

The eight drones are used in around ten projects related to forestry and wood. The drones vary in size, from small units weighing around one kilo to larger ones that require two people to load into a vehicle. A large drone can carry a heavy sensor or several smaller ones at the same time.

– Each sensor is unique and gathers specific data. If I can use several sensors during a single flight, it saves a lot of time when we’re collecting data.

Drones offer many advantages

All drones weighing 250 grams or more require a flyer ID. Many forest owners can benefit from simpler drones to quickly identify where windthrows have occurred after a storm or where a fire has started. The type of drone used by researchers and commercial operators is significantly more advanced.

For Annette as a researcher, drones offer many advantages. They are easy to manoeuvre, they can fly at much lower altitudes than an aeroplane, and they run on rechargeable batteries instead of fuel. It is also possible to programme exactly where and at what speed a drone should fly.

– Above all, drones allow us researchers to collect large amounts of data over extended periods of time. It’s easy to gather data high up in the forest canopy, where early changes in tree health often become visible. Thanks to the sensors, we’ll be able to make much more reliable and stable predictions about the forest.

Absolutely amazing tools

The sensors make it easier for researchers to carry out the same type of experiment a large number of times to investigate possible correlations. Annette, who wants to detect early signs of tree stress, can fly over the forest at very frequent intervals to identify changes in tree health.

– These are absolutely amazing tools. I think we’ll always need to complement drone flights with going out into the forest to see how things look on the ground – but we’ll have a hard time managing future forest research without drones and sensors.

Annette always begins her studies in a lab environment, where she can control external factors such as temperature and light. When explaining her work in the lab to others, she can – for now – only say that she is “subjecting small pine seedlings to stress in various ways”. Using the sensors in the lab, she examines how the pine seedlings change over time and can repeat experiments a large number of times.

The next step involves processing the data collected from the sensors, whether it comes from the lab or the forest.

– One single study can result in several thousand images, so we’re dealing with enormous amounts of data. We use AI tools to filter, organise, and classify the data. I dare say that a drone-pilot doing research spends the least amount of time on actually flying the drone.

A so-called multispectral camera reads the light reflected from the trees, visualising photosynthesis. This is compared against various indices and shows whether the trees are stressed by something – drought, nutrient deficiency, insects...

Various thermal images show the temperature of the trees, which can be useful in situations such as forest fires.