computer vision for close-to-ground drone flight operations in forestry

Important results from the project

The purpose is to create conditions for the commercialization of several very promising applications for drones in forestry. What they have in common is that, to varying degrees, they require autonomous computer vision and navigation of drones close to the ground. After an extensive analysis, it can be concluded that the perfect solution for Nordluft is currently not possible. In the near term, Nordluft will instead use various limited solutions in the different applications. Implementation has begun through tests with radar and lidar sensors as well as planned more advanced systems.

Expected long term effects

The conditions for commercialization of the various applications for drones in forestry are now, after the project, significantly improved. The result is a partially implemented plan for autonomous computer vision and navigation within a variety of applications. What has not already been verified in prototype tests is planned in detail and will be completed within the next 12 months. Within each application, there is now a clarity of what can be offered to the customer in the near term and what requires greater resource investment to be realized.

Approach and implementation

The project was carried out in two phases. Part 1 consisted of a feasibility study where available technology and solutions were studied, several meetings took place with suppliers to create a better practical understanding of capacity and pricing. The information picture was then compared with the company´s needs within the various applications. Results became an implementation plan for a range of inherently limited solutions, as a perfect solution was not deemed financially possible. In part 2, sensors began to be implemented and tests to be conducted. This work is ongoing.