Media

In DigiForest we are committed to the dissemination of our work. To achieve this, we are working to ensure that the project is being presented on public media and also advertising our work in social media! In this page, we facilitate the access to the media that different partners have generated throughout the project.

Online Tree Reconstruction and Forest Inventory on Mobile Robotic Systems

Terrestrial laser scanning (TLS) is the standard technique used to create accurate point clouds for digital forest inventories. However, the measurement process is demanding, requiring up to two days per hectare for data collection. We present a real-time mapping and analysis system that enables online generation of forest inventories using mobile laser scanners that can be mounted on mobile robots. Given incrementally created and locally accurate submaps—data payloads—our approach extracts tree candidates using a custom, Voronoi-inspired clustering algorithm. Further, we explicitly incorporate the incremental nature of the data collection by consistently updating the database using a pose graph LiDAR SLAM system. This enables us to refine our estimates of the tree traits if an area is revisited later during a mission.

Vision-based MAVs for forest

TU Munich has developed an MAV system that is capable of navigating through under-canopy environments while relying on passive visual sensing. This allows to reduce the overall payload of the MAV (enabling longer flight-times) while also reducing the overall cost of the system. The system can effectively navigate outdoor scenarios without colliding with the environment.

Building forest inventories with the AnyMal

Legged robots are increasingly being adopted in industries such as oil, gas, mining, nuclear, and agriculture. However, new challenges exist when moving into natural, less-structured environments, such as forestry applications. At DigiForest, we have developed a prototype system for autonomous, under-canopy forest inventory with legged platforms. We introduce a system architecture which enabled a quadruped platform to autonomously navigate and map forest plots. Our solution involves a complete navigation stack for state estimation, mission planning, and tree detection and trait estimation. The findings of this project are presented as five lessons and challenges. Particularly, we discuss the maturity of hardware development, state estimation limitations, open problems in forest navigation, future avenues for robotic forest inventory, and more general challenges to assess autonomous systems. By sharing these lessons and challenges, we offer insight and new directions for future research on legged robots, navigation systems, and applications in natural environments.

Reconstructing trees in the forest with a hand-held sensor box

University of Bonn has developed a full-system that allows to perform geometric and photometric reconstruction of forests. This enables to have the best texture and geometric information of our natural environments!

Anymal in the Forest

University of Oxford demonstrates how the Anymal can be deployed in Britain’s woodlands and how automation can help measure trees and forests.

Tree Instance Segmentation and Traits Estimation for Forestry Environments

To obtain per-tree KPIs of a forest, we need to identify all the tree instances present in our sensed data. To address this issue, University of Bonn has developed a method to segment individual trees and estimate the Diameter at Breast Height.

LiDAR place recognition in the forest

Many LiDAR place recognition systems have been developed and tested specifically for urban driving scenarios while environments such as forests and woodlands have been studied less closely. In DigiForest we analyzed the capabilities of four different LiDAR place recognition systems, using LiDAR data collected with a handheld device and legged robot within dense forest environments. We then incorporated the best performing approach, Logg3dNet, into a full 6-DoF pose estimation system—introducing and demonstrated the performance of our methods in three operational modes: online SLAM, offline multi-mission SLAM map merging, and relocalization into a prior map.

DigiForest: A mission

At DigiForest, we aim to make forestry more efficient by combining hardware with smart algorithms to enable foresters to perform data-driven decisions in an efficient fashion. This video gives a gentle introduction of the DigiForest project.