Use of LSTM Networks to Identify “Queenlessness” in Honeybee Hives from Audio Signals

Abstract

Honeybees are of vital importance to both agriculture and ecology, but honeybee populations have been in serious decline over recent years. The queen bee is of crucial importance to the success of a colony. In this paper, we contribute to addressing these problems by employing Long Short-Term Memory (LSTM), Multi-Layer Perceptron (MLP) Neural Networks and Logistic Regression approaches applied to audio data recorded from "queen-absent" and "queen-present" hives to provide a method of prompt detection of a hive lacking a queen. The initial resultsparticularly from the LSTM -are highly encouraging.

Buzzing with Intelligence: Current Issues in Apiculture and the Role of Artificial Intelligence (AI) to Tackle It

“…Detection of queenlessness in beehives [63]. Forecasting sudden drops of temperature in pre-overwintering honeybee colonies [64].…”

Prediction of Honeybee Swarms Using Audio Signals and Convolutional Neural Networks

“…A Mel spectrogram is a representation of an audio signal obtained by using the Mel scale frequency. The spectrum from the STFT described above is warped along its frequency axis f (in Hz) into the Mel-scale using triangular overlapping windows [21] using the formula: (7) where f denotes the normal frequency in Hz, and fmel denotes the corresponding Mel frequency [22]. The resultant Mel frequencies are then filtered using the formula below.…”

“…In recent years, acoustic measurements [4,5] have supported claims that experienced beekeepers could tell whether a hive lacked a healthy queen from changes in the sounds made by the bees. These insights have been used by researchers who used machine learning approaches to analyze acoustic signals from beehives to classify them as having a healthy queen present or not [6,7]. Swarms are also important for the wider success of a bee colony.…”

The Importance of Context Awareness in Acoustics-Based Automated Beehive Monitoring

“…More recently, with advances in sensing hardware, cloud computing, and machine learning tools, automated beehive monitoring tools have emerged to overcome these limitations, thus creating the field of precision beekeeping [2]. For example, systems based on beehive acoustics [3][4][5][6], hive weight [7][8][9], internal temperature [10,11], and humidity [11], as well as CO 2 monitors [11] or multisensor approaches [7,8,11], have been introduced.…”

“…To this end, the acoustic monitoring of beehives has emerged and is gaining popularity. In a recent literature review of beehive acoustics monitoring [24], several systems were reported showcasing tools for (i) bee activity detection [4,[25][26][27], (ii) beehive strength monitoring [28], (iii) queen absence detection [3,5,[29][30][31], (iv) swarming detection [22,23,[32][33][34], (v) pathogen or parasite infestation detection [35], (vi) detection of environmental pollutants and chemicals [36][37][38][39], and (vii) measuring of honeybee reaction to smoke [40], as well as overall beehive monitoring (e.g., identifying normal and abnormal hive, swarming duration, bee activity time) [41,42]. From a geographical perspective, most contributions have come from the USA, UK, Japan, Slovenia, and Italy.…”