Processing of multi-modal environmental signals recorded from a "smart" beehive

Hunter, Gordon, Howard, Donald, Gauvreau, Sylvain, Duran, Olga and Busquets, Rosa (2019) Processing of multi-modal environmental signals recorded from a "smart" beehive. Proceedings of the Institute of Acoustics, 41(1), pp. 339-348. ISSN (print) 1478-6095

Abstract

Environmental factors, including air pollution, noise, and decline in biodiversity, have become issues of major concern over recent decades. Air pollution and other environmental contaminants (such as pesticides) have led to concerns relating to the health and well-being of human, animal and plant populations, whilst changes in temperature and rainfall patterns raise issues of possible rises in sea levels, coastal erosion and changes to sustainable plant and animal populations. For example, the population of bees has experienced a marked decline in many countries, which is likely to have very serious consequences for agriculture and other plant life. Bees could also be sensitive to other environmental factors such as pollution, and our recent and present work is a first step towards monitoring bees for obtaining information from the wider environment. In this paper, we discuss the analysis and processing of multi-modal (sound, temperature, humidity, natural light level and air quality) signals recorded over several months from a sensor system of our own design. This sensor system was originally planned and constructed to monitor the health and well-being of honeybees in a beehive. However, we noted that same sensor system could additionally provide useful information concerning the local natural environment – for example, variations in air quality over time. We apply various signal processing methodologies both to individual signals and to the relationships between them, and discern some interesting patterns within the signals, including some relating to interactions between the environment and the activities of people living and working in the area. This work shows how a relatively simple and low-cost sensor system can be used to perform monitoring of the local environment, with a view to improving or preserving its quality, or at least limiting damage to it due to human interventions. Our sensor network (with Raspberry Pi microcomputer) cost approximately GBP £ 100 per system, or approximately GBP £ 200 per unit if audio and video recording, plus additional local data storage, were required. This should make the system reasonably affordable to farmers or environmental NGOs (e.g) in developing countries, for whom commercially-produced environmental monitoring systems may be too expensive.

Actions (Repository Editors)

Item Control Page Item Control Page