About the VETfarm Project

This course is designed to provide vocational training in Remote Sensing Image Analysis, specifically for Precision Farming. It is part of the VETfarm project, an Erasmus+ Capacity Building initiative aimed at establishing a "Demonstration Partnership for Pistachio Farming in Armenia: A Geospatial Approach." The course builds upon the Remote Sensing and GIS courses created within the VETfarm project framework, which serve as prerequisites. It emphasises the development of both theoretical knowledge and practical skills.

The primary objective of the VETfarm project is to build capacities in using integrated remote sensing and Geographic Information System (GIS) data for monitoring pistachio orchards and applying precision farming techniques.

The course focuses on using integrated remote sensing and GIS data to monitor pistachio orchards, including tree health, soil conditions, and water needs. Students will learn to analyse satellite imagery, integrate data from IoT sensors, and apply cloud-based tools to apply Management Zones (MZs) for optimised agricultural inputs, to improve efficiency and sustainability in farming practices.

The course focuses on building capacity to utilise integrated remote sensing (RS) and Geographic Information System (GIS) data for precision farming, with a specific focus on monitoring pistachio orchards and trees. It emphasises the analysis, modelling, and enrichment of Earth Observation (EO) data through IoT sensor networks to optimise agricultural inputs and monitor crop conditions.

A significant emphasis in this integrated course is on Knowledge Management in Pistachio Orchards. This course is dedicated to disseminating knowledge and promoting collaborative learning. It introduces participants to incorporating movable sensor platforms and digital tools in managing pistachio knowledge through RS and geoinformation systems. A key distinguishing characteristic of this course is its direct connection to the demonstration setup in Armenia. Participants will have the opportunity to manage a pistachio farm, thereby gaining practical and interactive training locally.

These courses will be linked to the equipment, staff and students training, and internship programs developed within the VETfarm project, promoting practical applications and collaboration. After identifying the context and updated needs of growers—considering relevant factors such as pests and management practices—the course will develop infrastructure, knowledge, and digital tools. These will evaluate the impact of pistachio farm management practices on tree/fruit growth and health using Mobile Labs. This closed-loop measurement, model updating, and action require skilled personnel.

Management Zones for Precision Farming: A Remote Sensing Approach

Concept: MZs are defined as "homogeneous production zones" within a field or orchard, identified based on similar characteristics like "yield-limiting factors."

Delineation: The course covers "delineating management zones (MZ) in pistachio fields using Sentinel-2 satellite data, aiming to optimise management practices." This involves analysing satellite and GIS data to identify specific regions with homogeneous characteristics, typically categorising them into "High yield," "Medium yield," and "Low yield" zones.

The generation of MZ is based on the WI tools and methods in information extraction from satellite data (see previous RS and GIS courses). For this, special attention is given to free, open-source systems for viewing, interpreting, and generating MZ.

Sampling Strategy: MZs are crucial for guiding sampling efforts. "Samples should always be taken from homogeneous areas or segments (MZ) to ensure reliable statistical results." A general guideline suggests dividing fields into three segments and taking at least three samples from each, totalling nine samples.

Application: MZs enable "spatially variable application of inputs," such as water and sampling efforts, leading to more efficient resource allocation and reduced waste.

Note:

• Management zones in the context of an orchard are defined in three dimensions plus time (3D+Time). These zones are influenced by the "resolution" of the "actor machines." To establish homogeneous regions, quad tree and voxel tree software can be employed. However, since trees are arranged on a planting grid, it is necessary to adapt the segmentation criteria from a "continuous" model to a grid-based approach.