Skip to main content

Autonomous mobile robot with computer vision

FOR THE GRADING OF TREES AND TREATED WOOD


ITCL (Castile and León Institute of Technology)

Castile and León Regional Government

A standalone air-to-ground solution that combines a drone with RTK and a quadruped robot equipped with LiDAR, 2D/3D cameras, thermal imaging, and computer vision.

What is it and what is it used for?

A standalone air-to-ground solution that combines a drone with RTK and a quadruped robot equipped with LiDAR, 2D/3D cameras, thermal imaging, and computer vision. It allows users to plan missions via a web interface, capture georeferenced data in forests and sawmills, detect stacks of logs, count timber resources, estimate volumes, and classify conditions such as treated wood, resulting in a unified and auditable inventory.

TECHNICAL SPECIFICATIONS
Leading organization
ITCL Technology Center
Administration
Castile and León Regional Government
Coordinator
Adrián Salazar (Head of the Robotics Technologies R&D Unit at ITCL)
Technologies
AI, IoT/Sensors, Satellites/Drones, Modeling, Autonomous Robotics, ROS2/Nav2, LiDAR, RTK, Photogrammetry, GNSS, Air-Ground Fusion, Web Interface
Target audience
Forestry professionals, government agencies, companies, forest owners, forest managers, and sawmills
Project framework
RetechFOR · Next Generation EU
4. VA Treated Wood

The problem it solves


It resolves the reliance on manual or remote-operated counts in the field, which are slow, scattered, and difficult to replicate. The lack of georeferencing and comparable data between forests and plantations leads to inventory discrepancies, poor visibility of stock by species, diameter, and condition, occupational hazards at logging sites, and forestry decisions based on incomplete information.

"Autonomous roboticsand computer vision to transform manual counts into georeferenced, reliable, and up-to-date forest inventories."

What makes it innovative?

Integration of Aerial and Ground-Based Measurements


The innovation focuses on integrating aerial and ground-based measurements onto a common geospatial framework: the drone generates 2D/3D maps, and the robot scans stacks at low altitude using advanced sensors. The processing of air-to-ground data enables the creation of an inventory organized by stack or lot—including coordinates, volume, count, and condition—accessible via a web platform.

Expected impact

It reduces the time spent on recounting and idle time, improves planning for logging, transportation, and timber utilization, and enhances safety by limiting manual patrols in rugged terrain. For government agencies, it provides spatial historical data for monitoring, conservation, and prevention; and for the industry, it offers an up-to-date inventory for purchasing, sales, and logistics.