Abstract
Precision agriculture requires information about crops and their environment to guidemanagement decisions. In this context, quadruped robots represent a research alternative foracquiring data from a ground-level, near-plant perspective. The objective of this work is toanalyze the potential applications and research opportunities of the Unitree Go2 robot inagricultural environments, considering its technical specifications, limitations, and adaptationrequirements. An exploratory literature review and a comparative technical analysis arepresented, linking background research in agricultural robotics, manufacturer specifications, andproductive task requirements. The information is organized using a matrix of applications,instrumentation, potential outputs, and validation criteria. Likewise, the possibilities of ground-based observation are conceptually compared with aerial photogrammetry and LiDAR. Theanalysis identifies four areas of opportunity: crop monitoring and characterization, environmentaland soil characterization, irrigation and infrastructure inspection, and spatial informationgeneration. The LiDAR and camera serve as key resources to investigate plant detection, rowidentification, dimension estimation, and incidence documentation. However, phytosanitarydiagnosis and specific soil measurements require additional instrumentation and procedures. Themain challenges include energy autonomy, stability during data acquisition, sensor integration,occlusions, localization, and measurement repeatability. It is proposed to complement aerialsurveys with targeted ground routes toward areas of interest. It is concluded that the agriculturalrelevance of the Go2 must be determined based on equipment configuration, operatingconditions, and the agronomic utility of the data. The study provides a research agenda withoutpresenting experimental validation and proposes progressive evaluations—including referencemeasurements and comparisons against wheeled platforms and manual procedures—to establishits advantages, constraints, and application conditions.