Raspberry Pi project aims to count your bees

Beekeepers no longer have to guess at the strength of their hives.

Raspberry Pi project, using photos to monitor the strength of bee colonies

A new Raspberry Pi project will capture activity at a hive's entrance in 10-second bursts. (Photo: Ratikova/Shutterstock)

When it comes to determining the strength of a particular honey hive, beekeepers often resort to estimating based on the tried-and-true method of eyeballing activity at the colony's entrance. For programmer Mat Kelcey, used to dealing in absolutes, this wasn't good enough.

"The first thing I thought when we set up our bee hive was 'I wonder how you could count the number of bees coming and going?' After a little research i discovered it seems no one has a good non-intrusive system for doing it yet," he wrote. "It can apparently be useful for all sorts of hive health checking."

Kelcey is definitely right about hive health corresponding to the strength of the colony. In spring, when warm temperatures beckon bees to begin flying again and the queen to start laying, the hive population can soar from thousands to over a hundred thousand by late summer. If by mid-summer the hive doesn't have consistently more activity than previous months, complications may be present — from a queen's failing health to disease. Gauging the population can be a useful indicator for inspiring action or simply leaving well enough alone.

To more accurately count bees flying in and out of a hive — a task near-impossible on warm sunny days when the term "busy as a bee" is come to life — Kelcey rigged up a Raspberry Pi, a standard pi camera and a solar panel to the underside of a top-bar hive. The Raspberry Pi is an incredibly small computer that, in addition to be a hacker's delight for all sorts of creative projects, can easily be unobtrusively setup to monitor hive activity.

Kelcey's system uses a camera to capture one photo every 10 seconds from 6 a.m. to 9 p.m. for a total of 5,000 photos a day. A software program then examines each image, ignores background noise, and marks each bee. Over time, with a bit of human-supervised tweaking, the network is able to train itself to better distinguish bees from other objects in the scene.

The data output is not only visually beautiful — you can see it over on Kelcey's blog — but also provides a daily snapshot of overall hive health. "I love how they all get busy and race home around 4 p.m.," Kelcey added.

Beekeepers interested in taking advantage of Kelcey's bee counter, or building upon it with their own hacks, can download the full source code over on GitHub.

Related on On Events:

purple flower Thymus serpyllum
honey hunter

MNN

Get the latest news delivered to your inbox

Follow us on social media networks