Articles
GUIDO.IO an autonomous UGV for horticultural greenhouse applications
Article number
1427_25
Pages
185 – 188
Language
English
Abstract
Unmanned Ground Vehicles (UGVs) are revolutionizing greenhouse applications, offering precise and efficient solutions for various tasks.
Equipped with advanced RGB and multispectral sensors, UGVs can navigate autonomously through greenhouse environments, in tight spaces between rows of crops, minimizing damage to plants while maximizing productivity.
These devices can operate continuously collecting data for mapping purposes or real-time task deployment ensuring sustainable agricultural practices.
A flexible UGV for greenhouse applications was developed under the project AGRIDIGIT (DM 36503.7305.2018 del 20/12/2018 – sub project AGROFILIERE) a.
The vehicle is characterized by an innovative flexible structure that allows variable working height and track width.
The autonomous navigation system is based on an advanced data fusion algorithm relying on ultra-wide band antennas and an RTK system, allowing precise navigation (around 4 cm) even in environments where GPS signal is poor.
The system onboard is developed using ROS 2 (robot operating system). The vehicle is equipped with a specifically developed hyperspectral linear sensor that through an artificial neural network (ANN) is able to recognize pathogens on fresh-cut salads and produce real-time output triggering signal to pilot a control system actuator.
The ANN can be trained on needs to recognize different pathogens and or weed in several crops.
Equipped with advanced RGB and multispectral sensors, UGVs can navigate autonomously through greenhouse environments, in tight spaces between rows of crops, minimizing damage to plants while maximizing productivity.
These devices can operate continuously collecting data for mapping purposes or real-time task deployment ensuring sustainable agricultural practices.
A flexible UGV for greenhouse applications was developed under the project AGRIDIGIT (DM 36503.7305.2018 del 20/12/2018 – sub project AGROFILIERE) a.
The vehicle is characterized by an innovative flexible structure that allows variable working height and track width.
The autonomous navigation system is based on an advanced data fusion algorithm relying on ultra-wide band antennas and an RTK system, allowing precise navigation (around 4 cm) even in environments where GPS signal is poor.
The system onboard is developed using ROS 2 (robot operating system). The vehicle is equipped with a specifically developed hyperspectral linear sensor that through an artificial neural network (ANN) is able to recognize pathogens on fresh-cut salads and produce real-time output triggering signal to pilot a control system actuator.
The ANN can be trained on needs to recognize different pathogens and or weed in several crops.
Publication
Authors
S. Figorilli, G. Manca, L. Bianchi, S. Mattogno, R. Di Ruzza, S.M. Vuth, D. Abruzzese, D. Carnevale, C. Carletti, L. Moscovini, C. Costa, C. Pane, F. Pallottino
Keywords
AI, ROS 2, rover, robot, remote sensing, image analysis
Online Articles (30)