SBIR Phase II: Machine Learning Driven Synthetic Sensor for Plant Water Stress

该SBIR二期项目将推进一种新型系统，用于感知、预测和处理植物胁迫，为永久和特种作物农民提供经济实惠的指导。这一指导方针将同时减少灌溉用水和农业化学品的使用，同时提高作物产量和农民利润。该项目将开发易于使用的设备和软件，提供与已安装的灌溉系统和流程兼容的指导。这项技术的潜在市场在美国是30亿美元，在国际上是300亿美元。该系统减少灌溉用水和化学品使用的能力具有深远的社会效益。该系统将保护稀缺的淡水资源，允许人口增长和灌溉土地的增加。该系统将减少农业活动造成的水污染，解决美国每年2100亿美元的水污染问题；此外，它将提高农业产量，确保不断增长的人口有足够的粮食，同时减少农药和杀菌剂的使用。该SBIR二期项目将展示一个软件和分布式硬件系统，用于测量和预测减产的植物非生物和生物压力源，并向农民提供干预指导，以减轻这些压力源。为了在局部水平上预测植物胁迫，必须融合和分析许多数据输入，以创建一个综合传感器，估计植物水分胁迫，预测小气候条件，并执行局部植物病虫害建模。该项目将推动超低功耗传感器阵列、广域网、机器学习和大数据解释的人机界面设计等领域的发展。该项目将通过增加小气候和病虫害模型预测算法来扩展测试系统，以便对灌溉、肥料、杀菌剂、农药和生物刺激素应用进行精确的时间和位置干预，以减轻植物胁迫。该奖项反映了美国国家科学基金会的法定使命，并通过使用基金会的知识价值和更广泛的影响审查标准进行评估，被认为值得支持。

This SBIR Phase II project will advance a novel system for sensing, predicting, and addressing plant stress, offering affordable guidance for permanent and specialty crop farmers. This guidance will simultaneously reduce irrigation water and agricultural chemical applications, while also increasing crop yields and farmer profits. The project will develop easy-to-use equipment and software that delivers guidance compatible with installed irrigation systems and processes. The addressable market for this technology is $3 B in the United States and $30 B internationally. The system's ability to reduce the use of irrigation water and chemicals has a profound societal benefit. The system will conserve scarce fresh water resources, allowing for population growth and an increase in irrigated land. The system will reduce water pollution from agricultural activities, addressing a $210 B annual problem in the United States; furthermore, it will improve agricultural yields, ensuring adequate food for growing populations, with reduced use of pesticides and fungicides.This SBIR Phase II project will demonstrate a software and distributed hardware system for measuring and predicting yield-reducing plant abiotic and biotic stressors, and delivering intervention guidance to farmers to mitigate these stressors. In order to predict plant stresses at a localized level, many data feeds must be fused and analyzed to create a synthetic sensor estimating plant water stress, predicting microclimatic conditions, and performing localized plant disease and pest modeling. The project will advance the fields of ultra-low power sensor arrays, wide-area networking, machine learning, and human interface design for big data interpretation. This project will expand a test system by adding microclimate and disease and pest model prediction algorithms for precision temporal and location-targeted interventions of irrigation, fertilizer, fungicide, pesticide, and biostimulant applications to mitigate plant stress.This award reflects NSF's statutory mission and has been deemed worthy of support through evaluation using the Foundation's intellectual merit and broader impacts review criteria.