Abstract

Habitat of 80% of all living species, forests cover approximately 30% of all landmasses on the Planet. Between 1990 and 2020 Earth has lost 4.4% of its forest area, or 178.000 square kilometres, as stated in the United Nations Environment Programme. Monitoring this environment is a key step for the conservation of its ecosystem and to fight climate change.

Within this context, the aim of this project is to design and implement an autonomous robotic system capable of navigating into a forest to create a Digital Twin (DT) of the environment under the canopy. The proposed system is built on a sensorized mobile robot with an artificial intelligence (AI) based controller.

Compared to a human-led effort, a robotics-based approach promises to deliver more objective and repetitive measurements, while being less dangerous for personnel and less expensive. Optimization and integration with other autonomous monitoring strategies is straightforward: e.g., satellite climate monitoring, UAV (Unmanned Aerial Vehicle) surveying.

The proposed DT of the forest is a map of the environment that includes information about elements of the forest: each tree is recognized and tagged with a unique ID, its position, diameter, tree type, vitality, etc. is determined. This model is the base for a forest census, vital in the context of climate change, plant conservation and disaster prevention, e.g., fires, landslides. It helps in the management of the economy and the environment of a forest. It gives insights into the quantity and quality of the trees, and it allows for efficient planning of forestry operations (e.g., tree felling, logging).

The key elements of this project are:

• The development of an autonomous robotic system capable of monitoring a forestry environment by means of sensors, such as Light Detection and Ranging (LiDAR), Inertial Measurement Units (IMU), Global Navigation Satellite System (GNSS), stereo camera, as well as a possible manipulator for collecting samples and performing measurements on the ground.
• The implementation of a navigation controller for the unstructured forestry environment based on an AI decisional model and a Simultaneous Localization and Mapping (SLAM) algorithm.
• The creation of a virtual model of the forest, for simulating the robot during the development of the controller.
• The application of the robotic system to mapping and censing a real forestry scenario, to obtain information on the environment, to evaluate the effects of climate change and the risk of possible disasters.

Partenariato

• Università degli Studi di Trieste
• Libera Università di Bolzano
• Università degli Studi di Udine

Importo del progetto

Importo totale dell’intero progetto   Euro 214.966,00
Importo del progetto Uniud             Euro 68.800,00
Contributo MUR Uniud                   Euro 68.800,00
Co-Finanziamento Uniud               Euro 0,00

Durata

• Data avvio progetto: 28 settembre 2023
• Data conclusione progetto: 27 settembre 2025

Link

https://prin.mur.gov.it/Iniziative/Detail?key=FiJNdaCuA71Xq3jYMAuZeQ%3D%3D