05 May 2022

Forest Inventory with AI Provides Great Environmental Benefits

The company Conifer Vision, based in Skellefteå, has the goal set on the forest. With unique AI solutions they provide valuable data for forest management and administration. It is both profitable and good for the climate.

In order to nurture and manage their forest in the best possible way and to be able to meet legal requirements for e.g., sustainability and biodiversity, forest owners need to know as much as possible about their forest. How many pines or spruces are there? How tall are the trees, how dense are they and are there signs of fungal or insect infestation? The information then forms the basis for decisions about support planting, clearing, thinning and much more.

Picture analysis with orthophoto shows how plants grow and which tree species they are. (Photo: Conifer Vision).

Roger Öhlund from Skellefteå is the CEO of the image analysis company Conifer Vision AB. He says:

• Mapping of the forest is an activity that has long been conducted with conservative methods. It has been done manually with random samples of randomly selected parts of the land. A laborious and time-consuming work where the result often is both subjective and low-quantitative.

Roger Öhlund is CEO and founder of Conifer Vision in Skellefteå. (Poto: Conifer Vision).

- The inventory is a bottleneck for forest owners. There is simply no capacity to make the analysis that would be needed for optimal management, says Roger.

Roger comes from a family with close ties to forestry. He is himself a forest owner and has many years of experience in image analysis projects and previous companies in the drone industry. In 2019, he started Conifer Vision together with his long-time business partner Jonas Granström.

Self-Learning AI Solutions

Conifer Vision’s solution for the forest inventory became a digital image analysis system that uses artificial intelligence, AI. Autonomous drones fly over vast areas and collect large amounts of data, which are stored in the cloud. With the help of deep neutral networks and semi-monitored machine learning, the collected information is analysed. Forest owners the use the analyses as a basis for decision making to conduct more efficient, environmentally friendly, and more profitable forestry.

Autonomous drones are sent out to collect data from large forest areas. (Poto: Roger Öhlund).

- There are many advantages to our cloud-based systems, says Roger. Scalability allows us to analyse large data volumes. It is a strong competitive edge.

- The systems can be applied to many different types of forest stands. For example, we have done a study on eucalyptus, a fast-growing tree with large commercial values from a global perspective.

- We can also help different types of customers, Roger continues. Smaller forest owners can use a simple online system. Our solutions can also interact with the current management systems used at the forest companies. Open APIs enable companies’ computer programs to communicate with ours.

- But what makes our solutions really unique is how we combine AI with semi-supervised machine learning, Roger explains. In this way, our systems are self-learning. The more data that is entered and processed, the more they learn.

Big Climate Gains

Growing forest binds carbon dioxide. The gigantic forest areas around the world therefore play a crucial role in counteracting the climate change through their ability to absorb carbon dioxide from the atmosphere. In a report from 2020, the Swedish Farmers’ Association (LRF) presents an extensive list of measures that together would increase the forest production in Sweden by 20%. More education, research and efforts against pests are some examples from the report. Forestry would be more profitable but most importantly, it would provide a climate benefit of tens of millions of tonnes of carbon dioxide per year.

The potential environmental benefits of Conifer Vision’s self-learning system are thus effective.

- With the analyses we produce, you can know exactly where the damage attacks or storm damage are, Roger explains. Aimed efforts can be made instead of fixing larger forest areas.

- It takes less from the forest resources, causes less damage from driving, and reduces emissions.

- If the growth in accordance with LRF’s report is 20%, the forest’s uptake of carbon dioxide would probably increase proportionally, Roger says. Our solutions can contribute by giving forestry tools to for the right measures, at the right time and in the right place.

Research and Future

Through an EU-project, Conifer Vision and Luleå University of Technology have initiated a collaboration where the researchers want to learn how AI can improve the understanding and interpretation of information during forest inspections with drones. Conifer Vision makes the systems available, and the researchers conduct the tests.

- This is an important collaboration for us as an innovative technology company, says Roger. The dialogue with the research and the knowledge created is worth its weight in gold.

Another important part of the company’s development is the dialogue with forest owners.

- Sometimes companies want the overall analysis, Roger explains. Sometimes they want to do the compilation themselves. We have to understand their need to become even better at what we do

Conifer Vision is currently in an intensive investment phase where most major forest players have shown interest. The company initially wants to establish in Sweden, which together with Finland are the innovative frontrunners in forestry.

- But we also eventually want to expand outside the country, in nations with similar inventory needs as ours but with much larger forest volumes, Roger says. Actors from Canada and Brazil have already shown interest.

- Eventually, we hope to be able to hire more employees for the office in Skellefteå, Roger continues. The market is huge all over the world! I almost do not see any limit to how much our company can grow.