Machine learning to keep bee hives healthy

Students at École Polytechnique Fédérale de Lausanne have created a machine learning algorithm to automatically count Varroa mites in beehives. The algorithm could help beekeepers monitor bee health more quickly and accurately.

Currently, Varroa mite infestations are tracked by counting the number of mite bodies found on a board placed underneath the hive. However, this is time-consuming and often inaccurate, because the mites are tiny and other debris also lands on the boards.

View a video demonstrating the current process and the new algorithm here.

Read the full article here and read coverage in TechCrunch here. See also the video A Bee's Eye View of a Healthy Food System in the Foodsource video library.

Publication
18 Feb 2019
Image
Image: Max Pixel, Bee bees honey, CC0 Public Domain
Image: Max Pixel, Bee bees honey, CC0 Public Domain
Actors
Fodder Category
Region