Opening the Future with Insect DXCreating new possibilities for insect science

At the Laboratory of Insect DX, we use digital technology to help solve social issues involving insects and to advance insect science. Our research rests on two pillars: “Agricultural Pest DX” and “Biodiversity DX.”

In Agricultural Pest DX, our motto is “Even if you block 99%, pests will exploit the remaining 1%!” We target flies that mechanically transmit pathogens, such as highly pathogenic avian influenza virus in poultry and pathogens involved in cattle infectious diseases, and work to automate forecasting and monitoring of their outbreaks through deep-learning-based image recognition. Using high-accuracy object detection models, mainly YOLO, we aim to evaluate pest occurrence and control effects quickly and with little labor, and to advance integrated pest management (IPM).

In Biodiversity DX, we use DNA metabarcoding, camera traps and image-recognition AI to make biodiversity monitoring more efficient, targeting flower-visiting insects, sap-feeding insects, fireflies and more. We are also developing methods to scientifically evaluate the state and changes of biodiversity by visualizing interaction networks based on co-occurrence and predator–prey relationships, and by estimating potential distribution ranges with species distribution models (SDM).

In addition, we use Kyushu University’s collection of roughly three million insect specimens as a digital resource, working on 3D modeling of specimen images, automated species identification, and prediction of distribution shifts under climate change. Through the fusion of insect science and information science, we contribute to adding value to insect resources and to a sustainable agriculture and society.