SBIR Phase II: Autonomous harvesting, mapping, and forecasting for fresh produce through application of robotics, computer vision, and machine learning

这个小型企业创新研究（SBIR）第二阶段项目的更广泛的影响/商业潜力在机器人应用，精准农业，深度学习，环境传感和协作机器人领域具有重要意义。自动化收获使人们能够种植健康和可持续的食物，农业机器人的长期市场机会为150 + B美元。该项目将为大型“受控环境”农场开发机器人解决方案，如温室和室外多通道。这些方法更密集地使用资本和劳动力，但需要更少的水（约减少90%），化学品使用（约减少50-70%）和肥料使用（约减少50%）。机器人可以减少所需的劳动力，并在对环境影响较小的情况下实现有竞争力的操作。 这个小型企业创新研究（SBIR）第二阶段项目将通过解决在农场等高度动态，精度要求高的生物环境中运营时所面临的前沿问题来推进计算机视觉/机器学习和机器人控制领域。通过发展和结合计算机视觉前沿的方法，该项目将开发一种新的方法来跟踪特定水果的时空变化，因为它随着时间的推移而移动和变化，收集有关植物生命周期的前所未有的详细数据。该项目将使用工厂级数据库集成可视化数据，以测试和改进浆果成熟度的检测和建模。最后，该项目将把现场光谱分析纳入成熟度分类。该奖项反映了NSF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。

The broader impact/commercial potential of this Small Business Innovation Research (SBIR) Phase II project is significant in the areas of robotic applications, precision agriculture, deep learning, environment sensing, and co-robots. Automated harvesting enables the ability to grow healthy and sustainable food, and the long-term market opportunity for agricultural robotics is $150+ B. This project will develop a robotic solution for large “controlled environment” farms, such as glasshouses and outdoor polytunnels. These use capital and labor more intensively but require less water (~90% reduction), chemical use (~50-70% reduction), and fertilizer use (~50% reduction). Robots could reduce the required labor and enable competitive operations with lower impact on the environment. This Small Business Innovation Research (SBIR) Phase II project will advance the fields of computer vision/machine learning and robotics controls by solving frontier problems faced when operating in highly dynamic, precision-requiring biological environments like farms. By evolving and combining approaches from the forefront of computer vision, the project will develop a novel approach to temporospatial tracking of specific fruit as it moves and changes over time, gathering data of unprecedented detail on plant life cycle. Using that plant-level database, the project will integrate visual data to test and refine detection and modeling of berry ripeness. Lastly, the project will integrate in-field spectrometry into ripeness classification. These objectives will underpin a novel precision agriculture solution.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.