计算机科学
机器人
葡萄园
异常检测
遥感
无人机
运动规划
精准农业
实时计算
领域(数学)
地面站
任务(项目管理)
自主机器人
树遍历
人工智能
工作流程
工作单元
计算机视觉
能量(信号处理)
软件部署
多光谱图像
能源消耗
能量收集
移动机器人
六足动物
仪表(计算机编程)
无线传感器网络
模拟
机器人学
作者
Eliza Murray,Stella Palmer,Joonseo Jang
出处
期刊:
日期:2026-06-05
卷期号:6 (01): 192-205
摘要
The integration of autonomous robotic systems into precision viticulture represents a transformative shift in how vineyard health, yield estimation, and disease detection are managed. Traditional monitoring techniques rely heavily on manual labor or homogeneous robotic deployments, both of which suffer from significant limitations in scalability, energy efficiency, and data resolution. This paper presents a comprehensive framework for heterogeneous field-robot collaboration, specifically focusing on the synergistic deployment of Unmanned Aerial Vehicles and Unmanned Ground Vehicles equipped with anisotropic sensors. Unlike omnidirectional sensors, anisotropic sensors such as RGB-D cameras, directional multispectral imagers, and phased-array radars possess strict directional constraints, limited fields of view, and range-dependent resolution. This spatial dependency introduces profound challenges in coverage path planning and task allocation. We propose a decentralized collaborative architecture wherein aerial units conduct high-level topological mapping and anomaly detection, dynamically generating high-resolution inspection waypoints for ground units. The ground units subsequently navigate the complex under-canopy environment using a novel anisotropic coverage algorithm that maximizes sensor utilization while minimizing traversal energy. Through extensive simulated and real-world vineyard trials, the proposed heterogeneous framework demonstrates a substantial improvement in inspection efficiency, reducing overall energy consumption while maximizing canopy coverage. The results indicate that accounting for sensor anisotropy within a heterogeneous collaborative framework significantly bridges the gap between aerial macro-monitoring and ground-level micro-inspection.
科研通智能强力驱动
Strongly Powered by AbleSci AI