The global collapse of honey bee colonies around the world is causing widespread concern Now it's hoped big data and machine learning can help us understand why and help stop the decline In a world-first, researchers are eavesdropping on bees inside their hive - sending the data to the Oracle Cloud This colony in southern England is one of the first connected to the World Bee Project Hive Network with a series of remote sensors measuring temperature, rainfall, humidity and even its weight.
SIMON POTTS, PROFESSOR OF BIODIVERSITY, UNIVERSITY OF READING, SAYING: "What does a healthy honey bee hive look like?
It will have a signature in terms of the activity of the bees, the heat, the temperature and all those sort of things.
And an unwell hive will have a different signature and we can start to understand actually what constitutes that by looking across many, many hives in many, many different environments." A third of all honey bees in the US have died and England's bees are vanishing faster than anywhere else in Europe.
Loss of habitat, climate change; pesticides, pathogens and pests are blamed in much of the world - but not everywhere SIMON POTTS, PROFESSOR OF BIODIVERSITY, UNIVERSITY OF READING, SAYING: "So the real strength of this project is to give great knowledge in realtime to allow beekeepers to predict or get ahead of the game in being able to treat their bees in a way that will stop them going down the slope to being unwell.
And we cant really do that.
An experienced beekeeper will know the signs but many beekeepers maybe dont know all those signs and there can be beekeepers around the world that dont have access to that kind of knowledge so suddenly we have a chance to learn, to take this collective understanding from many, many hives and distil it into what are really the key indicators of health." Researchers hope the UK pilot scheme can be rolled out globally - generating huge amounts of data.
JOHN ABEL, PROJECT DIRECTOR, ORACLE CLOUD TECHNOLOGY, SAYING: "We want more and more data.
The better the data, the better the training which means at some point the human wont be deciding what the bee keeper needs to do with the data.
The machine will be learning itself.
And actually the volume we're looking at, the machine needs that volume of data to make sure its making the correct predictions.
We want quality of predictions as well as quantity of data to make those predictions." Ultimately, the World Bee Project Hive Network's goal is to share the knowledge in order to prevent the further decline of the world's honey bee populations.