Solving The Bee Crisis with Machine Learning
Without the natural pollination bees provide, global food supply would deplete so rapidly, the effects would be disastrous. They’re an essential part of our ecosystem, a part of a delicate tapestry that works to naturally pollinate our crops. According to the British Bee Keepers Association, 1 in 3 mouthfuls of food we eat depend on bees. So, it’s pretty important that we work to save these amazing creatures.
BBC Earth Unplugged tells us over 70% of our crops are pollinated by bees, with honeybees being the biggest contributors to this. Not to mention, they’re responsible for $30billion a year in crops for the economy. Put simply, life without bees would not be sustainable.
But, why are their populations declining so rapidly? And what is Big Data and Artificial Intelligence doing to stop numbers from further decline?
Bees are under threat from many things. One of the most common is the Varroa Destructor — a parasite that feeds on bees and has the capability of destroying entire colonies. As is the nature of most parasites, they reproduce quickly and measuring a mere 1mm in length, they’re extremely difficult for beekeepers to detect. This is where the Bee Scanning app comes in. The app uses computer vision to help beekeepers detect early signs of the dangerous Varroa pests in their colonies. Using machine learning and object recognition, the red mites stand out on the bees bodies and an algorithm detects them in the images taken by the beekeepers.
Whilst it would be ideal to see the decline come to a halt, RoboBee is working to aid pollination by using mini-drones called RoboBees. Sticky horsehair underneath the drone collects pollen particles as they fly and rub off onto the next flower. They are currently manually controlled, but the team has reported they’re developing autonomous drones using AI.
There have also been dramatic changes in natural landscapes, with the loss of habitat being to blame. Global warming has also disrupted pollination, with hibernation times and flowers blooming not always matching up. Monitoring bees movements in accordance with their environment, the Bee Smart device allows beekeepers to remotely monitor their hives. The device uses sensors to track colony activity, temperature and humidity in the hive, and even mating patterns. The data collected is sent to the beekeeper through the cloud, via Bee Smart who process and analyze it. The use of big data by beekeepers will allow for a more proactive approach to beekeeping, even remotely.
Bee populations are also in jeopardy from harmful pesticides called neonicotinoids, and fungicides. The use of these chemicals is having devastating effects on bees. With firm data backing the fatal effects of certain pesticides, we can look to going some way to opening a dialogue with the farmers, firms, and producers of these chemicals to seek out alternatives. Or, in other cases take action to ban these totally. The Global Initiative For Honeybee Health is working on Smart Sensors that are fitted onto bees like little backpacks. These sensors monitor and collect data on how the bees interact with their environment. This information is then processed to see how disease, diet, weather, pesticides, and pollution are affecting colonies. Antennas are installed on entry sections to beehives so that when bees come and go, the sensor backpacks send data back via radio. This gives researchers a better insight into modeling bee’s movements and noting changes in behaviors, such as their ability to pollinate.
As the world faces many challenges, start-ups like the ones above are investing in making the world a better place for everyone by using big data, machine learning, and artificial intelligence. It’s through the innovation of technology that our ecosystems and workings of the natural world can be conserved.
If you’re interested in further reading on the Bee crisis and ways you can help, these are a great starting point,