<![CDATA[This study aims to formulate fruit detachment dynamics as the core concept for designing a selective vibratory mechanical harvester for robusta coffee. Plant-level average fruit removal force (FRF) and fruit mass data from five coffee plants were analysed for three maturity classes identified by color: red, orange, and green. An empirical FRF model was built using maturity code and fruit mass as predictors, after which a dynamic force model based on forced vibration, Fd = mA(2?f)², was used to simulate detachment behavior under different operating conditions. Model validation employed goodness-of-fit statistics and leave-one-out cross-validation, while selective harvesting performance was predicted through maturity-specific detachment probability curves. Mean FRF increased consistently with decreasing maturity, from 3.532 N in red fruits to 5.966 N in orange fruits and 8.034 N in green fruits, whereas fruit mass varied only slightly among stages. The empirical model showed high predictive accuracy (R² = 0.997; RMSE = 0.095 N), indicating that maturity status dominated the FRF response. Simulation results demonstrated that, at a vibration amplitude of 3 mm, equivalent threshold frequencies were approximately 99.9 Hz for red fruits, 129.9 Hz for orange fruits, and 151.3 Hz for green fruits. An optimum selective operating window of 3.60-5.91 N, equivalent to 101.0-129.4 Hz, was identified for maximizing ripe-fruit recovery while maintaining high harvest purity.]]>
Copyrights © 2026
