Anomaly Detection in Beehives using Deep Recurrent Autoencoders

  title={Anomaly Detection in Beehives using Deep Recurrent Autoencoders},
  author={Padraig Davidson and Michael Steininger and Florian Lautenschlager and Konstantin Kobs and Anna Krause and Andreas Hotho},
Precision beekeeping allows to monitor bees' living conditions by equipping beehives with sensors. The data recorded by these hives can be analyzed by machine learning models to learn behavioral patterns of or search for unusual events in bee colonies. One typical target is the early detection of bee swarming as apiarists want to avoid this due to economical reasons. Advanced methods should be able to detect any other unusual or abnormal behavior arising from illness of bees or from technical… 

Figures and Tables from this paper

Anomaly Detection in Beehives: An Algorithm Comparison

This paper compares multiple machine learning models for anomaly detection and evaluates them for their applicability in the context of beehives, finding that the autoencoder is the best multi-purpose anomaly detector in comparison.

Unsupervised Anomaly Detection on Multisensory Data from Honey Bee Colonies

Based on data recorded during the citizen science project Bee Observer BOB, it is demonstrated that algorithms exploiting an ARIMA Model and Receiver Operating Characteristics in combination with an underlying multi agent system are well able to detect and classify anomalies in honey bee colonies.

Anomaly detection at the apiary: predicting state and swarming preparation activity of honey bee colonies using low-cost sensor technology

This work proposes an approach for an automated prediction of anomalies in honey bee colonies that focuses on swarming events as well as the colony’s health state using a low budget DIY sensor setup, and compares methods from the domain of time series clustering.


It is found that using common convolutional neural networks in a Siamese setting can strongly improve the ability to detect human speech in recordings obtained from beehives, resulting in a reliable technique adhering to privacy regulations.

Semi-unsupervised Learning for Time Series Classification

SuSL4TS, a deep generative Gaussian mixture model for semi-unsupervised learning, is presented, to classify time series data, which can detect sparsely labeled classes and identify emerging classes hidden in the data.

SEGBEE: Mobile Application for Honey Segmentation in Apiary Boards

The development of the application \textit{SegBee}, a computational tool that performs the segmentation in the apiary plates, where there is the presence of honey, in an accessible, fast and practical way is presented.

Detecção de Anomalias em Padrões Acústicos, de Temperatura e Umidade Sazonais para Abelhas Melíferas (Apis mellifera L.)

O Distúrbio do Colapso das Colônias (Colony Collapse Disorder, CCD) é um fenômeno associado ao desaparecimento repentino de abelhas melíferas em colônias manejadas. Registrado nos EUA desde 2006, o



A predictive algorithm for mitigate swarming bees through proactive monitoring via wireless sensor networks

A predictive algorithm based on pattern recognition techniques, able to detect the increase in temperature in the hive (microclimate) responsible for the typical stress of bees that culminates in swarming is developed.

Anomaly Detection Using Autoencoders with Nonlinear Dimensionality Reduction

It is demonstrated that autoencoders are able to detect subtle anomalies which linear PCA fails and can be useful as nonlinear techniques without complex computation as kernel PCA requires.

Anomaly Detection with Robust Deep Autoencoders

Novel extensions to deep autoencoders are demonstrated which not only maintain a deep autenkocoders' ability to discover high quality, non-linear features but can also eliminate outliers and noise without access to any clean training data.

The temperature increase at one position in the colony can predict honey bee swarming (Apis cerana)

There was a linear heating process before swarming initiated and a specific location with the most apparent temperature increase was identified, which lies at the bottom part of the bee space between the wall of the hive and the first frame, which was towards the nest entrance.

Multivariate Industrial Time Series with Cyber-Attack Simulation: Fault Detection Using an LSTM-based Predictive Data Model

An approach based on an LSTM neural network to monitor and detect faults in industrial multivariate time series data is adopted and a Modelica model of part of a real gasoil plant is created to validate the approach.

Unsupervised Anomaly Detection With LSTM Neural Networks

  • Tolga ErgenS. Kozat
  • Computer Science
    IEEE Transactions on Neural Networks and Learning Systems
  • 2020
This work jointly train and optimize the parameters of the LSTM architecture and the OC-SVM (or SVDD) algorithm using highly effective gradient and quadratic programming-based training methods and provides extensions of this unsupervised formulation to the semisupervised and fully supervised frameworks.

Multi-Sensor Prognostics using an Unsupervised Health Index based on LSTM Encoder-Decoder

This work proposes a Long Short Term Memory based Encoder-Decoder (LSTM-ED) scheme to obtain an unsupervised health index (HI) for a system using multi-sensor time-series data and evaluates its approach on publicly available Turbofan Engine and Milling Machine datasets.

Deep Learning for Anomaly Detection: A Survey

This survey presents a structured and comprehensive overview of research methods in deep learning-based anomaly detection, grouping state-of-the-art deep anomaly detection research techniques into different categories based on the underlying assumptions and approach adopted.