Acoustic signal processing systems for intelligent beehive monitoring

Abstract

Bees, as pollinators and producers of honey and medicinal products, play a crucial role in human life and environmental sustainability. Emerging Smart Beekeeping technologies utilise various methodologies in apiology, agricultural science, computer science, and electrical engineering. A significant part of these technologies includes data-driven and intelligent condition monitoring systems that can ideally imitate expert beekeepers. This paper shows that the acoustic signals generated by bees form an efficient and reliable source of knowledge about the beehive and its bee colony. Also, it proposes an acoustic signal processing system for intelligent and data-driven beehive monitoring. The proposed system includes acoustic data acquisition, noise reduction, feature extraction and machine learning techniques for inferential or predictive data analysis. This system can be used for different monitoring purposes; however, this paper focuses on queenless beehive identification. Finally, this paper reports a flexible experimental setup for developing and testing intelligent beehive monitoring systems.