b+WSN: Smart beehive with preliminary decision tree analysis for agriculture and honey bee health monitoring

A WSN for extended monitoring of beehive activity and condition has been developed.Data collected from a beehive were analysed from a multi-disciplinary perspective.A decision tree algorithm describing hive/colony status was proposed and evaluated.An algorithm for predicting short term rainfall local to the hive was also proposed.The algorithms were deployed in network with a minimal energy increase (5.35%). United Nations reports throughout recent years have stressed the growing constraint of food supply for Earth's growing human population. Honey bees are a vital part of the food chain as the most important pollinator for a wide range of crops. It is clear that protecting the population of honey bees worldwide, as well as enabling them to maximise their productivity, is an important concern. In this paper heterogeneous wireless sensor networks are utilised to collect data on a range of parameters from a beehive with the aim of accurately describing the internal conditions and colony activity. The parameters measured were: CO2, O2, pollutant gases, temperature, relative humidity, and acceleration. Weather data (sunshine, rain, and temperature) were also collected to provide an additional analysis dimension. Using a data set from a deployment at a field-deployed beehive, a biological analysis was undertaken to classify ten important hive states. This classification led to the development of a decision tree based classification algorithm which could describe the beehive using sensor network data with 95.38% accuracy. Finally, a correlation between meteorological conditions and beehive data was observed. This led to the development of an algorithm for predicting short term rain based on the parameters within the hive. Envisioned applications of this algorithm include agricultural and environmental monitoring for short term local forecasts (95.4% accuracy). Experimental results shows the low computational and energy overhead (5.35% increase in energy consumption) of the classification algorithm when deployed on one network node, which allows the node to be a self-sustainable intelligent device for smart bee hives.

[1]  Met Éireann.  Met ireann - The Irish Weather Service , 2018 .

[2]  J. Settele,et al.  Economic valuation of the vulnerability of world agriculture confronted with pollinator decline , 2009 .

[3]  J. Ross Quinlan,et al.  Induction of Decision Trees , 1986, Machine Learning.

[4]  Miguel Ángel Porta-Gándara,et al.  Automated Irrigation System Using a Wireless Sensor Network and GPRS Module , 2014, IEEE Transactions on Instrumentation and Measurement.

[5]  F. Ratnieks,et al.  Long-range foraging by the honey-bee, Apis mellifera L. , 2000 .

[6]  Geoff V. Merrett,et al.  Predicting discharge using a low complexity machine learning model , 2015, Comput. Electron. Agric..

[7]  Marcelo A. Aizen,et al.  The Global Stock of Domesticated Honey Bees Is Growing Slower Than Agricultural Demand for Pollination , 2009, Current Biology.

[8]  S. Ferraria,et al.  Monitoring of swarming sounds in bee hives for early detection of the swarming period , 2008 .

[9]  Antonio Criminisi,et al.  Decision Forests: A Unified Framework for Classification, Regression, Density Estimation, Manifold Learning and Semi-Supervised Learning , 2012, Found. Trends Comput. Graph. Vis..

[10]  Michele Magno,et al.  Towards persistent structural health monitoring through sustainable wireless sensor networks , 2011, 2011 Seventh International Conference on Intelligent Sensors, Sensor Networks and Information Processing.

[11]  Andrei Lucian,et al.  Original paper: Identification of the honey bee swarming process by analysing the time course of hive vibrations , 2011 .

[12]  Michael A. Brown,et al.  Testing a grassroots citizen science venture using open design, "the bee lab project" , 2014, CHI Extended Abstracts.

[13]  Vincent Bretagnolle,et al.  ECOBEE: a tool for long-term honey bee colony monitoring at the landscape scale in West European intensive agroecosystems , 2014 .

[14]  Michele Magno,et al.  Big brother for bees (3B) — Energy neutral platform for remote monitoring of beehive imagery and sound , 2015, 2015 6th International Workshop on Advances in Sensors and Interfaces (IWASI).

[15]  Mauro Serpelloni,et al.  Self-Powered Wireless Sensor for Air Temperature and Velocity Measurements With Energy Harvesting Capability , 2011, IEEE Transactions on Instrumentation and Measurement.

[16]  L. A. F. Heath,et al.  Carbon Dioxide Activation of Spores of the Chalkbrood Fungus Ascosphaera Apis , 1987 .

[17]  Michele Magno,et al.  b+WSN: Smart beehive for agriculture, environmental, and honey bee health monitoring — Preliminary results and analysis , 2015, 2015 IEEE Sensors Applications Symposium (SAS).

[18]  Joe-Air Jiang,et al.  An imaging system for monitoring the in-and-out activity of honey bees , 2012 .

[19]  Eva Crane,et al.  The World History of Beekeeping and Honey Hunting , 1999 .

[20]  JOHANNES GEHRKE,et al.  RainForest—A Framework for Fast Decision Tree Construction of Large Datasets , 1998, Data Mining and Knowledge Discovery.

[21]  Dongjian He,et al.  Decision support of farmland intelligent image processing based on multi-inference trees , 2015, Comput. Electron. Agric..

[22]  J. R. Quinlan.  Learning With Continuous Classes , 1992 .

[23]  Aleksejs Zacepins,et al.  Challenges in the development of Precision Beekeeping , 2015 .

[24]  Lior Rokach,et al.  Top-down induction of decision trees classifiers - a survey , 2005, IEEE Transactions on Systems, Man, and Cybernetics, Part C (Applications and Reviews).

[25]  Mark J. F. Brown,et al.  Brood pupation temperature affects the susceptibility of honeybees (Apis mellifera) to infestation by tracheal mites (Acarapis woodi) , 2005 .