Índice de curva ótima para implantação de lavouras cafeeiras

Abstract

Coffee cultivation is extremely important to the Brazilian agricultural economy, requiring the adoption of techniques that increase productivity and reduce operating costs. The use of highaccuracy elevation data is essential for the efficient planning of mechanized planting rows, particularly in mountainous regions such as southern Minas Gerais, Brazil. This study aimed to develop an Optimal Curve Index (OCI) for designing coffee planting rows by considering mechanization variables, topographic characteristics, and the optimization of the number of transplanted seedlings per hectare. The research involved analyzing mechanized areas, determining optimal curvature and slope parameters, and establishing practices to prevent plant damage caused by machinery traffic during agricultural operations. Technologies for acquiring topographic data were compared, including a LiDAR sensor mounted on a remotely piloted aircraft (DJI Matrice 350 RTK equipped with a Zenmuse L2) and GNSS receivers using RTK correction. LiDAR elevations were processed using the hgeoHNOR2020 geoid model and referenced to the SIRGAS 2000 / UTM Zone 23S coordinate system. The initial analysis indicated that the data were not normally distributed, as confirmed by the Shapiro–Wilk test (p = 2.79 × 10−21 for GNSS and p = 2.97×10−21 for LiDAR). The Wilcoxon test revealed a statistically significant difference between the measurements, with a mean bias of 1.56 m, indicating the need for vertical correction of the GNSS data. The OCI model, validated through Monte Carlo simulations, showed that tractor–implement combinations can negotiate tighter curves, whereas self-propelled harvesting imposes stricter constraints due to the geometry of the harvester. The harvesting-tunnel length adopted in the model was 3.66 m, resulting in an absolute critical radius of 6.61 m. Depending on the local slope, the harvester required minimum safe radii greater than 20 m to operate without mechanical collisions with coffee plant trunks, drastically reducing overall operational suitability in fields with sharp curves. It was concluded that the OCI, which can be supplied with data obtained affordably from sensors already installed on agricultural machinery, is a feasible tool for guiding planting-row design in the field and reducing the need for future physical interventions in coffee plantations