Abstract: The system classifies aspects of beehive phenomenon by fusing sensor data from inside a beehive and human collected data from the same beehive using machine learning algorithms. The human collected data labels the sensor data indicating various aspects of beehive states. By collecting observation data from people looking into the beehives equipped with sensors, the sensor data is augmented with information. The resulting sensor and labelled dataset are used to train machine learning algorithms to classify beehive health states using sensor data alone. This system is used for remote monitoring of honeybee hives using in-hive sensors.