Digital phenotyping and precision monitoring of honeybee colonies: emerging role of IoT and machine learning for predictive analysis

Abstract Predictive and observational monitoring of hives is essential for ensuring colony health, optimising pollination, and enhancing apicultural productivity. Traditional methods of hive monitoring, reliant on manual observation, are increasingly inadequate in addressing the complexities of modern beekeeping. Consequently, advancements in sensor technologies and information and communication technologies are revolutionising hive monitoring, offering unprecedented precision and scalability. The study proposes a predictive analysis of bees to clarify Internet of Things (IoT) and machine learning-assisted hive monitoring for reducing mortality, analysing their behaviour and comprehending their needs. The predictive monitoring utilises advanced acquisition tools, IoT, and data handling methods to examine physical and behavioural characteristics of bees. Our survey reveals that the implementation of IoT-based predictive monitoring can enhance comprehension and optimise best-practice management of beehives leading to a better understanding of environmental changes, which is crucial for adapting to diverse climates, improving pollination management and rapid response to hive health issues. The study also explored potential negative impacts of sensor usage on bees, as well as the commercial implications, economic aspects, challenges, and benefits of computer-aided automatic beehive monitoring systems. After carefully reviewing and analysing the information, several conclusions and potential areas for further exploration were identified.