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涉及蜂巢的健康和相关的每年150亿美元的农业服务。蜂巢容易受到寄生的狄斯瓦螨的影响;除非对蜂巢进行处理，否则瓦螨的存在最终会对任何蜂群造成致命影响。监测蜂箱中的螨水平需要部分拆卸蜂箱，最好的计数方法牺牲了一些蜜蜂。全国120，000多名养蜂人中的大多数人没有充分检测螨虫。该项目的预期产出是一种低成本、电池供电的“物联网”设备，可以放置在蜂箱内，被动地计数瓦螨种群，并将结果报告给养蜂人。计数可用于指示何时需要螨治疗以及治疗是否有效。它可以防止昂贵和不必要的预防性治疗，可能导致螨种群的耐药性。该项目培育健康和高产的蜂群，同时减少养蜂人的工作量，减少对蜂箱不必要的化学处理。该项目的智力价值在于使用近红外光波长的新颖研究，以确定蜜蜂和瓦螨外骨骼之间存在可接受的对比度，以及使用现代人工智能技术识别寄生瓦螨，该技术可以缩小用于边缘计算环境，该环境在功率可用性和处理速度方面具有显著限制。第一阶段的工作将集成一个红外相机与可变波长窄带红外LED照明，并拍摄无寄生虫和螨虫出没的蜜蜂。人类研究人员将识别螨虫，并使用评分图像训练人工智能引擎来检测和计数螨虫。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。

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.