BeeMind AI: Development of an Artificial Intelligence-Based System to Assess Honeybee Health, Behavior, and Nutrient Effects

This research proposed and built the first integrated AI-based honeybee health assessment system called BeeMind AI. The BeeMind AI system had eight sensors including a microphone, temperature and humidity, carbon dioxide, atmospheric pressure, and camera, which enabled BeeMind AI to monitor both in-hive and external conditions. BeeMind AI has several diverse applications due to its ability to analyze honeybee movement and behavioral patterns to determine honeybee health, and it was used to evaluate the effects of four nutrients on honeybee health through video analysis in two experimental settings, one in a newly designed tri-chambered maze based on a Delayed Matching-to-Sample procedure, and another in a free-flying homing paradigm. The free-flying experiment was conducted to study the effect of nutrients on return rates of honeybees at distances of 300 m, 500 m, and 800 m, and it was found that the base return rates of the control group even at 800 m was close to 75%. It was observed for the first time that C60 nanoparticles had significant positive effects on learning, memory, and flying capabilities, improving return rates by around 9% at 300 m, 16% at 500 m, and 20% at 800 m, while neonicotinoid pesticides had negative effects on return rates, reducing them significantly by up to 30%. The developed BeeMind AI system has a significant impact on honeybee-related research, especially in the evaluation of honeybee learning and memory.