Biodrone is specialised in drone services within the forestry and agricultural industry. Biodrone wanted to create a simple and accessible method for forest analysis to contribute to sustainable precision forestry. Sopra Steria contributed
to the project with a cloud-based solution where users can use artificial intelligence to replace manual processes. The solution opens new opportunities to run efficient and sustainable precision forestry.
The customer's situation
Biodrone is specialised in drone services within the forestry and agricultural industry. They deliver a range of sustainable and innovative solutions and take on assignments throughout the country. Their services include surveying, seeding, spraying and
fertilising from the air. Modern and sustainable forestry is about measuring and analysing the forest, so that we can utilise and take care of the resources in it in the most efficient and sustainable way possible. Biodrone wanted to create a simple
and accessible method for forest analysis using drones and artificial intelligence to contribute to sustainable precision forestry. This is done by creating a web portal that can be used by everyone in forestry, without them needing expertise in geographic
information (GIS), photogrammetry or artificial intelligence (AI).
Sopra Steria's contribution
The project uses state-of-the-art artificial intelligence technology from Amazon Web Services (AWS) to recognise forest and plant varieties from images taken with a drone from up to 200 meters high. Sopra Steria develops AWS infrastructure for training,
testing and validating machine learning models. The system also supports fine-tuning of the models, such as active learning, which is necessary when photos are taken in different light, season, angle and with a different camera sensor. In addition,
the customer can train their own models that can be used for mapping and analysing objects. The purpose of the portal is to offer forestry a simple and intuitive system that uses drone technology in forestry. It should be equally easy to use for everyone,
regardless of expertise. Sopra Steria has developed a Cloud-based solution where the customer can upload images or video, process data, and use artificial intelligence to replace otherwise manual processes. The customer gets several ways to
analyse the images and look at statistics in a map-based interface. The project uses state-of-the-art technology in artificial intelligence to recognise forest and plant varieties from images taken with a drone from up to 200 meters high. Sopra Steria
Team developed infrastructure for training, testing and validating machine learning models.
Value for the customer
With this solution, Biodrone’s customers can fly over the forest, take pictures and upload to the web portal. They receive a density map that describes the number per species of tree and the average height per species of tree. This provides significant
information and is much more accurate than the manual method used today. Forests require continuous management. Many incidents require measures that must be taken relatively quickly to ensure good forest health - a prerequisite for efficient and sustainable
forestry. It is therefore important to quickly map, identify and implement measures to reduce the consequences of tree falls, top breaks that can drive injuries. Windfall and top break must also be reported to insurance companies, and in Biodrone's
web portal you can easily get an overview of the changes. Driving damage is time-consuming to identify and measure, but has a major effect on future planting, drainage and the health of the forest. Unwanted plant species or disease in the plants can
have a major impact on the state of the forest if not detected early enough. Identification of these is time-consuming and, in many cases, completely random. Then it is often too late. This project enables Biodrone to offer an innovative service for
everyone in the forest industry, from forest owners, via cooperatives and contractors to management. The solution opens new opportunities to run efficient and sustainable precision forestry. Users can standardise reporting and documentation, which
is important in dialogue between different players. Today, Biodrone’s clients use the portal to analyse several hundred forest areas in Norway and Sweden. Biodrone has also secured an agreement for continuous mapping and analysis for Trøndelag
County Council, which means monitoring tens of thousands of hectares per year. Similar agreements are underway with Oslo and Viken. The world's largest drone manufacturer, DJI, has also partnered with Biodrone as its forestry partner. Together, they
will demonstrate how drones can be used to revolutionise forestry.
Sopra Steria shows unwavering professionalism in all parts of the project. Their professionals have a high level of competence and expertise. This ensures precision and attention to detail in the project. They have also shown an incredible ability to
adapt to new circumstances, both when it comes to changing project requirements or meeting unexpected challenges. Sopra Steria has a clear and effective communication style characterised by openness, cooperation and trust. All this makes it a pleasure
to work with them.
Benefits for Business users
|Forest owners can survey their own forest and ask for price offers from several actors or carry out the work themselves based on drone data. Forest owners now have the option of more accurate mapping of much larger areas. They can track development
over time to confirm the effect of the measures and get a better basis for securing the best price from several contractors.||Contractors use the portal as a work tool to uncover where measures are needed and calculate the price and time spent on young forest care. This gives a price that is calculated for more complementary information and therefore most competitive.
The contractors can document completed work and prove the effect.||Administration: Forest managers use the platform as a tool for documenting carried out rejuvenation and young forest care, and to increase their own efficiency in this work. This provides more control and monitoring of the measures, and more
accurate measures mean lower costs. By identifying unwanted plant species earlier, forest managers can take measures to prevent their spread. |