Adaptive AI-enabled and Context-enhanced Mobile Intelligence for Climate-smart Pest Management to Optimise Sustainable and Resilient Farming

在世界范围内，农作物受到无脊椎动物害虫的威胁，这些害虫会造成取食损害并传播植物病毒。高水平的虫害可造成高达80%的产量损失。目前，建议农民遵循经济门槛，并在超过门槛时进行管理干预。通常，阈值被定义为虫害水平，超过该水平，预计作物将遭受经济损害。随着对杀虫剂使用限制的增加和更多抗药性害虫种群的出现，种植者正在寻找更可持续的虫害综合管理实践。要有效地实施综合害虫管理实践，需要三个组成部分：1)准确识别当前的害虫(S)；2)关于阈值及其药效的准确信息；3)关于当地/地区害虫种群抗药性的信息。然而，有许多障碍限制了对IPM原则的理解：准确识别无脊椎动物有害生物很困难，需要分类学培训，这是种植者往往缺乏的技能；目前的阈值在田间条件下几乎没有得到测试和验证，限制了种植者的信心；关键害虫的抗药性信息在空间上有限，主要是在国家基础上提供的。在最近的一个项目中，我们开发了这个问题的早期解决方案，通过建立一个人工智能驱动的害虫检测模型来识别小麦作物中的害虫(创新项目10002902)。在此，我们建议在该项目成功的基础上，将人工智能驱动的害虫检测模型扩展到其他农田作物的害虫，并将更多信息整合到最终用户的输出中，以解决IPM吸收的其他障碍。为了实现这一目标，我们将把害虫检测模型扩展到油菜籽和马铃薯的地上害虫，整合关键害虫：甘蓝蚜虫、桃蚜-土豆蚜虫、土豆蚜虫和甘蓝茎跳甲的区域特定抗药性状况，并测试和验证这些害虫的子集的阈值。我们的主要产品将是一个智能应用程序，它提供害虫检测支持，突出显示已识别害虫的当前阈值，并提供关于区域害虫种群抗药性状况的信息。人工智能模型的开发将由谢菲尔德大学领导；害虫管理建议和阈值测试的提供将由ADAS领导；杀虫剂抗性测试将由利物浦大学领导；智能应用程序用户界面的开发将由Mutus Tech Ltd.领导。

Worldwide, crops are threatened by invertebrate pests which cause feeding damage and transmit plant viruses. High levels of infestation can cause up to 80% yield loss. Currently, farmers are advised to follow economic thresholds and to apply management interventions when thresholds are exceeded. Generally, thresholds are defined as the level of pest infestation above which it is expected the crop will suffer economic damage. As restrictions on insecticide use increase and a greater number of insecticide resistant pest populations emerge, growers are looking towards more sustainable integrated pest management (IPM) practices.To effectively deploy IPM practices three components are required: 1) accurate identification of the pest(s) present; 2) accurate information on thresholds and their efficacy; 3) information on insecticide resistance of the local/regional pest population. However, there are numerous barriers that restrict the uptake of IPM principles: Accurate identification of invertebrate pests is difficult and requires taxonomic training, a skill that growers often lack; current thresholds have received little testing and validation under field conditions, limiting grower confidence; and insecticide resistance information for key pests is spatially-limited and primarily provided on a national basis.The central barrier for IPM uptake is lack of grower confidence in their ability to identify a pest. In a recent project we developed an early-stage solution to this problem by building an AI-driven pest-detection model to identify insect pests in wheat crops (Innovate project 10002902). Here, we propose to build on the success of this project by expanding the AI-driven pest-detection model to pests of other arable crops and by integrating more information into the end-user output in order to address the other barriers to IPM uptake. To achieve this we will expand the pest detection model to above-ground pests of rapeseed and potato, integrate region-specific insecticide resistance status for key pests: cabbage aphid, peach-potato aphid, potato aphid, and the cabbage stem flea beetle, and test and validate thresholds for a subset of these pests.Our main output will be a smart-app that provides pest detection support, highlights the current threshold for the identified pest, and provides information on the insecticide resistant status of regional pest populations. AI-model development will be led by The University of Sheffield; provision of pest management advice and threshold testing will be led by ADAS; insecticide resistance testing will be led by The University of Liverpool; and the development of the smart-app user-interface will be led by Mutus Tech Ltd.