Conference Agenda

Manual planting of cassava is a tedious, time-consuming, and arduous operation, which is the most conventional production in the Philippines. A cassava mechanical planter was developed, hitched to a four-wheel tractor and designed to plant cassava horizontally. The planter parts include the hopper, cylindrical metering device, transmission system, delivery tube, furrow opener, and furrow closer. The planter’s performance was evaluated in terms of uniformity of planting, field efficiency, and percentage stake germination. A 2 x 3 complete factorial experimental design was used, and data were analyzed using Analysis of Variance (ANOVA) for the test of significance. The optimum planting condition using the Response Regression Method (RRM) was also identified. The effect of hopper load on the field efficiency and uniformity of planting is not significant. The obtained uniformity of planting at speeds of 3.0, 3.4, and 3.8 kph is 18.12, 28.52, and 41.58 cm, respectively, while the machine’s field efficiency at different speeds is 60.13, 57.52, and 54.84 %, respectively. The planted stakes had 100% germination. The optimum planting condition was achieved at full hopper load and speed of 3.0 kph. The developed planter benefits include labor reduction, uniformity in planting, and an increase in field capacity.

Uniform seed placement is critical for garlic yield and profitability, yet conventional garlic planters can produce miss and double-seedings that are difficult to identify during operation. This study develops a real-time garlic seed monitoring and mapping system for an eight-row-garlic planter operated in a bucket-type seed-metering device by integrating an RGB camera with a GNSS/INS receiver on an embedded Jetson TX2 platform. YOLO-based detectors based on a family of YOLOv8–12 were evaluated and a lightweight YOLO11n was then selected for Jetson TX2 deployment, thereby providing frames per second (fps) of 25.6 with high precision (95.3%) and recall (97.6%). To georeference seeding events, seeding positions were estimated using a calibrated longitudinal offset between the camera detection point and the actual seed discharge location. A five-bucket calibration test showed a mean deviation of 0.25 cm from the theoretical 100 cm offset. The in-row spacing estimates achieved a mean absolute error (MAE) of 4.11 cm, a root mean square error (RMSE) of 5.12 cm, and a coefficient of determination of 0.98 compared to relative to tape-measured ground-truth (GT). Field testing across 384 seeding events yielded F1-scores of 0.97 for single seeding and 0.95 for missing seeding, with an overall classification accuracy of 96.6%.

Mechanical damage during harvesting and handling is a major contributor to overall grain loss. The objective of this study was to develop mechanics-based models capable of accurately predicting grain damage resulting from mechanical handling processes. A discrete element method (DEM) simulation was developed to predict the impact damage of maize kernels in lab-scale testers and was validated using an industrial-scale handling system for maize and soybeans. At the lab-scale, the mean root-mean-square error between the simulation outputs and experimental results was 0.05, indicating the strong predictive accuracy of the DEM models. In addition, the damage model provided insight into the locations within the device where kernel damage occurred. Industrial-scale experiments showed that grain damage increased with increasing impeller speed and decreasing feed rates in a rethresher system, which was consistent with the model predictions. Under similar test conditions, the handling process caused more damage to maize than to soybeans. Sensitivity analysis of the simulation results showed that the predicted damage fractions were strongly affected by DEM input parameters, particularly particle shape, in tested systems. Overall, the developed damage models have strong potential to guide the design and operation of grain handling processes to minimize kernel damage and improve grain quality.

The Discrete Element Method (DEM) is critical in agricultural machinery design due to its accuracy in predicting granular material behavior. This study compares real-world tests of twin-disc fertilizer spreaders, conducted according to the ASAE 341 standard, with simulations performed using Altair EDEM software. The research examined four drag models (Haider-Levenspiel, Morsi-Alexander, Schiller-Naumann, and Ganser) and two lift models (Magnus and Saffman) using a full factorial experimental design. To represent the airflow generated by the discs, a perpendicular airflow was applied in the simulations, and the resulting distribution data were statistically analyzed. Univariate analyses demonstrated that all drag and lift models, including their interactions, significantly influenced the fertilizer distribution. According to RMSE and MAE error analyses, the Schiller and Naumann drag model yielded results closest to the experimental data. Consequently, the use of the Schiller and Naumann model is recommended for R&D studies involving DEM simulations of twin-disc fertilizer spreaders.

This study investigated the effects of changing the material feeding position on the machine performance of the hammer mill, which is widely used in mixed feed production. The surface of the crushing chamber, featuring a circular cross-section material entry opening, was designed to alter the material feeding position of a domestically produced horizontal rotor-type hammer mill. With this designed surface, the distance of the feeding position from the rotor shaft was varied, and the angle of the feeding position relative to the sieve symmetry axis (α, °) was adjusted at eight levels for each distance (0°, 45°, 90°, 135°, 180°, 225°, 270°, 315°). After producing the design, the geometric mean diameter (mm), increase in surface area (mm² g⁻¹), specific energy consumption per mass (kWh kg⁻¹), and specific energy consumption per surface area (kWh m⁻²) resulting from the hammer mill’s processing of wheat were determined. As a result of this research, the material feeding position that maximizes energy efficiency was identified for the first time in the literature.

The production of Cipolla Bianca di Margherita PGI seeds requires dedicated mechanization strategies capable of ensuring high seed integrity during threshing. In the absence of a standardized mechanical design methodology for onion umbel processing, this study presents the conceptual development and preliminary structural validation of a rasp-bar-based thresher specifically designed for seed extraction. A three-dimensional model of the threshing drum and rasp-bar assembly was developed and analyzed through finite element analysis (FEA). Static structural simulations were performed by applying distributed loads representative of operational resistance forces generated during threshing. This simulation-driven approach enabled optimization of geometric parameters and mass distribution without relying on iterative empirical prototyping, thereby reducing development time and enhancing design robustness. Particular attention was dedicated to drum geometry, rasp-bar configuration, concave clearance, and peripheral speed to achieve an effective balance between impact and shear forces, while minimizing the risk of seed damage. The FEA procedure allowed identification of stress concentration areas and evaluation of stress distribution patterns to verify compliance with material strength limits and adequate safety margins under combined loading conditions. The proposed methodology offers a replicable framework for the preliminary engineering of specialized threshing systems supporting the modernization of PGI onion seed production.