Abstract Purpose Efficient orchard spraying requires uniform canopy coverage while reducing pesticide drift and application rates. To achieve this, variable rate applications (VRA) are preferred. However, most current studies on VRA focus on directing the spray liquid to the tree canopy. To address these challenges, a real-time variable-rate air-assisted orchard sprayer was developed, integrating laser sensors and a hydraulic-driven turbofan controlled by LabVIEW software. Methods The objective of this study was to develop a real-time variable-rate orchard sprayer capable of controlling the outgoing airflow as well as the spray liquid based on canopy characteristics. Prior to testing, the sprayer was adapted to each orchard, and its performance was evaluated through deposition and drift trials under different canopy geometries and leaf densities. Results With VRA, air velocities were maintained between 3 and 5 m s⁻¹ at the canopy’s outer edge, a range critical for achieving adequate air capacity and droplet transport within tree canopies. VRA left 25% of the conventional application (CA) tracer on leaves, yet this value was considered adequate when the spray index data was analyzed. VRA also achieved drift reductions in ground losses (86.95%) and airborne drift (89.98%). Conclusion The system’s trigger mechanism, guided by real-time canopy data from laser sensors, proved adaptable to diverse orchard conditions. Thanks to the sprayer, more successful spraying with less drift was achieved with less pesticide usage (69.90%) and less fuel consumption (14.78%). These findings highlight the potential of this precision agriculture technology to enhance spraying efficiency, conserve resources, and minimise environmental impact.

The paper presents an approach to generating routes for agricultural machinery based on digital field maps using QGIS software and satellite navigation technologies. The proposed algorithm includes the following steps: preparing input data, converting field boundaries to a metric projection, creating a rectangular grid with a step equal to the working width of the implement, generating cell centroids, assigning them ordered route indices, and building a continuous trajectory of machine passes. To improve practical usability, line geometry smoothing reduces the number of sharp turns. The resulting routes are exported to formats compatible with GNSS terminals and autopilot systems. The route-generation method relies on open-source software, does not require specialized commercial platforms, and can be integrated into precision farming workflows to reduce overlaps, optimize fuel use, and lower the technogenic impact on soil. The approach can serve as a practical tool for agricultural enterprises and as a teaching example for training specialists in GNSS technologies in the agricultural sector.

Abstract Purpose Hyperspectral sensing (remote or proximal) has emerged as a pivotal tool to classify plant materials (seeds, leaves, and whole plants), pharmaceutical products, food items, and many other objects. Thus, hyperspectral sensing is one of the most frequently used technologies in research articles published by this journal, and it was therefore found relevant to address two methodological issues, which (based on Google Scholar searches) appear to be over-looked or ignored in >94% of hyperspectral sensing studies: 1) the "small N, large P" problem, when number of spectral bands (explanatory variables, “P”) surpasses number of observations, (“N”) leading to potential model over-fitting, and 2) absence of independent validation data in performance assessments of classification models. Methods Based on simulations of randomly generated data, risks associated with these issues were illustrated. This communication explores and discusses consequences of over-fitting and risks of misleadingly high accuracy that can result from having a large number of variables relative to observations. Moreover, connections of these issues with radiometric repeatability (levels of stochastic noise) are highlighted. A method is proposed wherein a theoretical dataset is generated to mirror the structure of an actual dataset, with the classification of this theoretical dataset serving as a reference. Conclusion By shedding light on important and common experimental design issues, the principal aim is to enhance methodological rigor and transparency in classifications of hyperspectral sensing data and foster improved and effective applications across various science domains, including precision agriculture.

Abstract Purpose Precision agriculture requires detailed knowledge of the within-field variation of yield forming factors and the productivity potential of each area of the field. The goal of this work was to use a case study to test all the steps of the process of creating a crop model-based yield map from soil mobile soil sensors and determine the impact of uncertainties and inaccuracies on the results. Methods Soil texture maps (0- 90 cm of depth) of a field were derived from mobile sensors and used as input for the process based deterministic crop growth model HERMES to produce a high-resolution yield map.Results Compared to actual yield maps, the simulated yield map successfully identified the major differences in productivity within the field, although some spatial variation was lost during the simulation, mostly at the point of translating soil texture maps into soil water retention parameters. The model also showed a tendency to overestimate yield across the entire field. A crop model simulation based on measured soil parameters resulted in a yield prediction accuracy of about 10% higher than a simulation based on estimated (mapped) soil parameters. Conclusion The loss of spatial variability, although measurable, occurred at a scale that might not have a significant impact on the site-specific management plan. Most yield map inaccuracies can be attributed more to model calibration than to the mapping process itself.

Abstract Purpose Effectively transferring knowledge of digital innovations, such as precision agriculture, is critical for farmers to make adoption decisions to meet current and future farming challenges. Stronger knowledge flows from Extension systems to stakeholders benefit adoption of precision agriculture. Methods This study used a three-round Delphi technique with a panel consisting of 15 precision agriculture experts who are academic extension faculty at land-grant universities in the Southeastern United States to identify the key knowledge, skills, and abilities needed by Extension agents to facilitate effective precision agriculture interactions. Results A total of 13 key competencies were identified. The findings revealed that the majority category identified in the 13 key competencies was the skill component, followed by the knowledge component, indicating the need to expand Extension agents' skills and knowledge related to precision agriculture. Conclusion The top key competency of basic precision agriculture awareness underscores the importance of ensuring Extension agents are aware of precision agriculture applications in their region. While all items were considered relatively important to the panelists based on the rate data, sustainability awareness, and equipment operation skill competencies were ranked as a lower priority, indicating Extension agents do not necessarily need to understand how to use the precision agriculture technology or manage the data but should function as change agents who introduce farmers to precision agriculture and connect them with relevant experts. The findings inform future tailored professional development programs and educational resources for equipping Extension agents with the needed competencies and guide future assessment of precision agriculture competencies among Extension agents.