Integrated Proximal Sensing of Soil and Crop

该研究项目的最终目标是推进近端土壤和植物传感技术，使低成本获取作物生长条件时空变化的准确信息成为可能。近端传感系统依赖于从与目标接触或距离目标不到2米的传感器收集信号。了解和理解土壤和植物特性的异质性有助于农民和其他土地管理者优化决策过程，以开发有利可图、可持续和环保的业务。在2017-2022年期间，重点将放在将几种土壤传感技术集成到一个单一的现场土壤分析仪中，并开发一种优化的方法，以便在试点和自主操作模式下部署该系统。最小的集成封装将包括：几个离子选择膜（pH，硝酸盐，钾，钠等），数码显微镜，分光镜，土壤呼吸传感器以及机械阻力和含水量传感器的集成盒。该系统将在矿物和有机土壤条件下进行测试。集成的植物传感平台将扩展到结合多光谱、超声波和热测量与低等级激光扫描和数字图像组件。该平台将作为作物表型的基本单位；它将在商品作物（玉米和大豆）以及饲料和蔬菜作物上进行测试。最后，将开发一种基于云的地理空间数据分析工具，将使用正在开发的系统获得的近端遥感数据与传统的高密度数据层（如田间地形、土壤表观电导率、产量图、卫星和低空图像等）混合在一起。邻域搜索聚类过程将用于通过最小化成本和最大化目标专题地图的准确性来增加信息的价值，这些地图揭示了空间可变的土壤生产力和潜在的作物胁迫评估。该项目将在未来五年内至少聘用6个hqp。研究结果将在精准农业科学界、相关农业企业和服务提供商以及作物生产者协会中传播。

The ultimate goal of this research program is to advance proximal soil and plant sensing technologies to enable low-cost acquisition of accurate information about spatial and temporal changes in crop growing conditions across a landscape. Proximal sensing systems rely on gathering signals from transducers placed in contact, or less than 2 m away from the target. Knowing and understanding the heterogeneity of soil and plant properties help farmers and other land managers optimize their decision-making process to develop profitable, sustainable, and environmentally friendly operations. During the 2017-2022 period, the emphasis will be placed on integrating several soil sensing techniques in a single, on-the-spot soil analyzer and the development of an optimized methodology for deployment of this system in both piloted and autonomous modes of operation. The minimum integrated package will include: an integrated cartridge of several ion-selective membranes (pH, nitrate, potassium, sodium, etc), digital microscope, spectroscope, soil respiration sensor as well as mechanical resistance and water content sensors. The system will be tested in both mineral and organic soil conditions. The integrated plant sensing platform will be extended to combine multispectral, ultrasonic and thermal measurements with low-grade laser scanning and digital imagery components. This platform will serve as the basic crop phenotyping unit; it will be tested on commodity crops (corn and soybean) as well as forage and vegetable crops. Finally, a cloud-based geospatial data analytic tool will be developed to blend proximal sensing data obtained using systems being developed with conventional high-density data layers, such as field topography, apparent soil electrical conductivity, yield maps, satellite and low-altitude imagery, etc. The neighborhood search clustering process will be used to increase the value of information obtained by minimizing the cost and maximizing the accuracy of targeted thematic maps revealing spatially variable soil productivity and potential crop stress assessments. The program will engage at least six HQPs over the next five years. The results will be disseminated within the precision agriculture scientific community, relevant agribusinesses and service providers as well as crop producer associations.