STTR Phase I: Automated Continuous Varroa Mite Monitor for Honeybee Hive Health

这个小企业技术转让第一阶段项目旨在解决蜜蜂蜂箱的健康问题，以及每年150亿美元的相关农业服务。蜜蜂蜂箱对寄生的瓦罗亚粉尘很敏感；除非蜂箱得到治疗，否则瓦罗粉尘的存在最终会证明对它所侵袭的任何蜜蜂群体都是致命的。监测蜂箱中的尘螨水平需要对蜂箱进行部分拆解，而最好的计数方法是牺牲一些蜜蜂。全国12万多名养蜂人中的大多数都没有进行充分的螨虫检测。该项目的预期产出是一种低成本、由电池供电的“物联网”设备，它可以放置在蜂箱内，被动地计算瓦罗亚螨虫的数量，并将结果报告给养蜂人。该计数可以用来指示什么时候需要治疗粉尘，以及治疗是否有效。它可以防止昂贵和不必要的预防性治疗，这些治疗可能会导致螨类种群产生抗药性。该项目培育了健康和多产的蜂群，同时减少了养蜂人的工作量，并减少了对蜂箱进行不必要的化学处理。该项目的智力优势在于，利用近红外光波长来确定蜜蜂和瓦罗阿螨的外骨骼之间存在可接受的对比度的新颖研究，以及使用现代人工智能技术识别寄生的瓦罗阿螨，这些技术可以缩小规模，用于在功率可用性和处理速度方面具有重大限制的边缘计算环境中。第一阶段的工作将整合一台带有可变波长窄带红外LED照明的红外相机，并拍摄未寄生虫和蠕虫出没的蜜蜂。人类研究人员将识别螨类，并使用评分图像来训练人工智能引擎来检测和计数螨类。这一奖项反映了NSF的法定使命，并通过使用基金会的智力优势和更广泛的影响审查标准进行评估，被认为值得支持。

This Small Business Technology Transfer Phase I project addresses the health of honeybee hives and the associated $15 billion in annual agricultural services. Honeybee hives are susceptible to the parasitic Varroa destructor mite; unless the hive is treated, the presence of the Varroa mite eventually proves fatal to any bee colony it infests. Monitoring mite levels in beehives requires partial disassembly of the hive and the best counting method sacrifices some bees. Most of the nation’s 120,000+ beekeepers do not adequately test for mites. The expected output of this project is a low-cost, battery-operated “Internet-of-Things” device that can be placed inside hives, passively counting the Varroa mite population, and reporting the results to the beekeeper. The count can be used to indicate when mite treatment is needed and whether the treatment was effective. It prevents costly and unneeded prophylactic treatments that can lead to resistance in mite populations. This project fosters bee colonies that are healthy and productive, while reducing beekeeper workload and reducing unnecessary chemical treatments for beehives. The intellectual merit of this project is in the novel investigation of the use of near-infrared light wavelengths to determine where acceptable contrast exists between the exoskeletons of honeybees and Varroa mites, and the recognition of parasitic Varroa mites using modern artificial intelligence techniques that can be scaled to down for use in an edge computing environment that has significant limitations in terms of power availability and processing speed. The Phase I effort will integrate an infrared camera with variable wavelength narrowband infrared LED illumination, and photograph unparasitized and mite-infested honeybees. Human researchers will identify mites and use scored images to train an artificial intelligence engine to detect and count the mites.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.