Smart beehive monitoring system for identification of relevant beehive events

Instead of a human environment, this paper deals with an intelligent environment in a context of a beehive. Precision beekeeping is a concept that is gaining prominence in the apicultural community as it relies on data analysis from a series of smart sensors. Here, we used smart beehive monitoring system with weight time series analysis to recognize different characteristic events in a beehive, which is helpful to beekeeper to reduce the number of unnecessary interventions. Using data obtained in a pilot study, and synthetic data, we developed an interpretative module for the events in a hive. We obtained a 0.86 accuracy and 0.87 F1 for rule-based approach and up to 0.89 accuracy and 0.88 F1 score for a machine-learning approach.