An AI-Based Digital Scanner for Varroa destructor Detection in Beekeeping

Area of Science:

• Agricultural Science
• Entomology
• Artificial Intelligence

Background:

• Honey bees are vital for agriculture and environmental health through pollination.
• Honey bee populations face significant threats from parasites like Varroa destructor mites, which transmit viruses and cause colony losses.
• Current methods for detecting Varroa mites rely on time-consuming and often inaccurate human visual inspection.

Purpose of the Study:

• To introduce and evaluate a novel digital portable scanner coupled with an AI algorithm (BeeVS) for detecting Varroa mites.
• To assess the accuracy, reliability, and speed of the BeeVS device compared to conventional human visual inspection methods.

Main Methods:

• A digital portable scanner and AI algorithm (BeeVS) were developed for Varroa mite detection.
• The device analyzes images of sticky sheets placed under beehives to count naturally fallen mites.
• The scanner was tested over 17 weeks, analyzing weekly samples from 5 beehives, and compared against human visual counts.

Main Results:

• The BeeVS device demonstrated high measurement repeatability (R² ≥ 0.998).
• For sheets with at least 10 mites, BeeVS achieved a cumulative percentage error below 1%, significantly outperforming human visual observation (approx. 20% error).
• The digital scanner offers a substantially faster monitoring capability for multiple beehives compared to manual counting.

Conclusions:

• The BeeVS device is a highly accurate and reliable tool for Varroa mite detection.
• Its speed and accuracy make it a valuable asset for beekeepers and scientists in monitoring honey bee health.
• This AI-driven approach provides a significant advancement over traditional methods for managing Varroa mite infestations.