You see lots of numbers and facts bandied about with regards to bee hives and winter. Temperatures, how the cluster moves, when they cluster up, etc. But you don’t actually see the data. I know that the centre of a winter cluster of bees is supposed to be approximately 90F all the time, but only because I’ve read it somewhere. I’ve never measured it, nor do I have any idea how it was measured in the first place, when it was done and what the winter conditions were like. That’s all about to change.
Greg and I have been talking about wiring up a hive with sensors for at least three years. It’s relatively straightforward: you just have to figure out how you want to arrange them, how may you need, how they’ll be connected and how will you be collecting the data? Simple.
The mechanics of it all are pretty easy. Maxim International makes a very affordable temperature sensor ($1 apiece in bulk on eBay) that needs no power supply, is accurate to with 0.5°C, is dead simple to hook up and you don’t need specialised or expensive software to query. So, I ordered 30 of ’em. 😎
That was two years ago. This past Saturday, I finally got around to hooking the works up. It was raining, after all.
First up was soldering leads on to them. All 30 of them. I miss my closeup vision. The sensor itself is about 2cm long. Add about 30cm of wire and some heat-shrink tubing and you have a nifty little package ready to go. Well, almost. I wanted to put the sensors about 10cm down from the top of the box, so each one is threaded through a piece of polyethylene tubing and the whole works was stuffed through a hole in the inner cover into the hive.
It took all morning Saturday to wire them up. Good thing it was raining outside.
Sunday dawned beautiful and sunny: a perfect day to shove electronics into a ball of stinging insects. It was a good day to die, so to speak. 😉
We’re trying to do this as scientifically as we can, so next step was to mark the location of the sensors and drill the holes into the inner cover. What I wound up doing was a 5×5 grid (less one sensor for the opening in the inner cover). Columns are between frames: one right down the centre and two spaces on each side. The wholes are spaced 10cm apart in the column.
As long as you get the alignment of the frames right, it’s all good. I had to loosen them up a bit to get the sensors in. LOL
OK, time to annoy the bees. Swap inner covers, jam the sensors in and wire them together.
They didn’t care. Mostly. What I’d like to know is how do the bees know the second you open your veil? They were totally relaxed, so I took my veil down. Almost immediately, two fly into my face and tried to kill me. Sheesh. Veil went back on and they went away.
There are 24 in the hive, one in the air space above the inner cover and another hanging out the back of the hive for an external temperature. They’re all wired together in parallel and connected to 150 feet of network cable run through the long grass back to the house. That cable gets connected to a an electronic converter and plugged into an old laptop running Linux Mint.
I’m using an open source program called DigiTemp to poll the sensors and store the output in a simple text file. Much to my absolute amazement, it worked the first time I checked it. I was polling every 10 minutes, but 24 data entries six times per hour winds up being more than 44000 lines of data per day. Yeah, no.
I’ve started learning things already, too. The cluster temperature itself is very uniform throughout. It ranges from 28 – 32C. The temperatures are very stable: less than 2 degrees fluctuation over 24 hours. I’ve dropped the data polling back to twice per hour and if the temperatures wind up staying this stable, I may drop it back again.
As we learn more and start to visualize the data, we’ll post it here.