Save the Bees Beefore It’s Too Late

This weekend I attended a NSF Big Data conference where lots of PIs presented their various research projects. There were lots of interesting projects, ranging from practical real world experiments to theoretical algorithm designs.

The most interesting one I found was about tracking bees as they leave and enter beehives and using that data to determine various things about their behavior. This behavior can in turn be used by ecologists to to protect the honeybee population, which has been dropping for a number of years.

The aspect of this research that most fascinated me was the finding that the bee’s entrance/exist habits can be grouped into three main categories: time specific, specific to broad, and broad. Time specific bees tend to leave and enter the hive at approximately the same time every day, while the bees classified as “broad” tend to leave and enter at random times. Furthermore, the bees classified as “specific to broad” start off leaving the hive at predictable times early in their life, and become less predictable as they age.

They also took plenty of images of bees and used image recognition to determine how much pollen they are carrying.

One of the researchers stated that something he wanted to improve in the future was study individual hives and see how hives differ. The current data is a mash-up of over a dozen hives.

We also discussed how tagging the bees to track their behavior can actually have an affect on the bees behavior as well. The researcher noted that bees marked with green paint started disappearing altogether. Turns out, the green paint made it easier for birds to identify the bee and eat them.

Overall, I think this project is very interesting and it is vital for saving the bees.