 Right. So, I'd like to have the pleasure of introducing Joe Meadows here. I'm a geologist, so I want to say Geo Meadows, but apparently it's Joe. Yeah, it's Joe. It's Joe from Kittimenser, who is going to talk about Beehive Monitor based on a Raspberry Pi. So, I'll turn it over to you, Jill. Let's give him a very warm welcome. Thank you. We've had a little trouble with the display, so I'm struggling with the mouse points. I think I've got the hang of it now. We're on extended screen instead of duplicate, which throws me a bit. Okay. Raspberry Pi, obviously, Honey Pi is a nice name, and we thank the Beatles for the intro. Any Pi will do. I have two of these monitors working at the moment. I hope to show you one shortly, live off the internet. So, one of them is using a zero. Another one is using a two. Three is a bit of a waste, really. There's not that much going on in there, but we'll come on to that. So, let's rapidly progress. Okay. So, the Honey Pi is now on a Raspberry Pi. It started off on an Arduino, and I've got an Arduino I can show you. Anyone wants to have a look later, but I moved on for that because I wanted to add video and audio, and that's a lot easier with a Raspberry Pi as it's doing remote software updates. So, Raspberry Pi it is. Okay. We've covered that. Why bother? Mainly because it's fun and it's interesting to see what the bees are up to. 24, 7, 12, no matter the weather. You don't have to go down and look at them and weigh things and all that. You can be anywhere in the world and you'll get the information you need. Beekeepers have managed for thousands of years without Honey Pi, but they've had some nasty surprises when they checked the bees after being away a while. So, this is the advantage of the remote monitoring. It gets you ahead of the game. For example, colonies die in the winter. As soon as this happens, you will know and you will get the indication that they are in trouble, so you can take avoidable action when they are running out of food because we are weighing the hive all the time. With a bit of knowledge, you know where the danger zones are. Okay. You can also detect the fact that the colony is swarmed and you can do something about that. For example, I was in holiday in Pembrokeshire a couple of years ago. I saw one night I made the mistake of going on the internet to see what my bees were doing, which my wife told me never to do. I saw they'd swarmed, so I convinced her that we had to come back the next day. I was not popular. But the next day, sure enough, I caught the swarm because when they swarm... How many beekeepers are here? Okay. You know about the swarming situation. Yeah. When they swarm, they congregate locally. So, you've got about a day to catch them normally and then they sit there and send all the scouts out trying to find a nice home. And then they decide where to go. It could be a couple of miles away and then you've lost them. But you've got a window of opportunity. So, if you can see they've swarmed, you can do something about it. You can also get in the hive and the beekeepers will understand this. Deal with all the queens. I won't go there. It's to beekeeper stuff. You can also see in the spring, if you're leaving the honey on too long, again, certain types of honey, you need to get out the hive quickly. You can also get an indication of whether the hive's lost a queen. I need to do something about that. So, there's lots of preventative stuff you can do. Okay. This is one of my hives. It's polystyrene stack. You're not seeing the point of what I'm seeing. I'll just see if I can use this. Is that showing through the screen? Yeah. Great. Okay. Well, this is the brood area where the eggs are laid. And there's a soup as well. We try to get the bees to put the honey. So, okay. Now, this is the load cell. And if we just go back to the previous one, what I meant to show you was that on the front of this hive, just about here, there's a hinge. And just about here, there's a hinge. So, it's a cancel lever. All the weight's on the back of the hive. And that's where the load cell sits. So, the load cell screws to the base and takes the weight of the top. So, it sees half the weight of the hive. But that's okay. We've got software. We can cope. So, yeah, that's the load cell. Left-hand side would be clamped to the base. And the right-hand side floats and takes the weight of the hive. All right. So, we can use any pie. So, this offers a low-cost monitoring solution. For about 70 pounds, you can build this. Very rough figures. 10 pound for a pie. 30 pound for the load cell. 5 pound for the temperature sensors. 10 pound for the interface modules. We'll look at them shortly. 10 pound for the box. And the hardware. Well, if you've got a garage like mine full of stuff, it didn't cost you anything. Maybe allow 5 pound. The Bosch load cell, it took me a long time to home in on this. It's rated at 40 kilogram, but because the only putting half the weight on it, it can cope with 80 kilogram total hive weight, which I had back in July. All the information here, by the way, is on a website called honeypie.me.uk. And the last slide will give you all the links you need to get the information should you wish to pursue it. Okay. Here's the circuit diagram. Raspberry Pi. Any Raspberry Pi will have the GPIO here. Standard connector. There's a module required to interface to the load cell. The load cell is a Wheatstone bridge arrangement. You get very, very small signal. You need decent differential amplifiers. So it's all built into this module. You don't have to worry about it. Very simple wiring. The analog interface. There's a four-way analog input. So I've just got two sensors. It obviously could be extended. I've home in on LM34s. They're nice, easy ones to use. Just five volts ground and a signal. So, couldn't be simpler. Okay, I think what I'm gonna try and do now, if I can deal with the extended screen problem is show you something from today. I might need a bit of help here. So this setup's not the normal one for me. I want to get onto the internet. That's the one. Yeah, great. Okay. So this is real time. All the way from Kiderman store. Right. Okay, let's see what the Bs are up to today. I can't see that on my screen. Yeah. So, what have we got? Yeah, yeah, yeah. The top left is a day of weight on the hive. What we're seeing is the weights going up a little bit. I suspect by the end of the day, it won't have gone up very much because if you look on the bottom left curve, you've got the weight over 10 days. And forget the drop at the beginning. That's me taking off some supers. But it's generally a downward trend. The Bs are actually going out and expending more fuel in the way of honey than they're compensating for by what they bring back. So it's a downward trip now. It's typical this time of year. They're just consolidating, planning for the winter as bees do. Top right is the brood temperature. All right. It's between 32.5 and 34.5. It's pretty good. It's showing me that there's brood in there. There are eggs, there are larvae. Otherwise, it wouldn't be so high. They wouldn't need to keep it so high. Bees have to keep it at that temperature to maintain the larvae. The outside temperature is the bottom right. Okay. So last night, it wasn't too cold. So there's an attempt. We'll know that. Right. I need to get back on. So I need to do control. That. And that. Excellent. No, it's not that. Oh, great. Okay. Back on PowerPoint. Yeah. So we've had a look at the circuit diagram. We've had a look at today's information from Kiddo Minster. Nothing of great note. Right. The software, very briefly, there's a Raspberry Pi operating system, free download, obviously. The pihive.py is a Python script that I've done available. That deals with the load cell via the HX711 module. And it deals with the two temperature sensors. And it passes data to thingspeak, which is the magic thing about it. We'll come back to that in a second. There's another module which is save zero. This is for getting the zero, the tarring on the system, so that when there's no ways on the load cell, you measure the zero and then the software pulls that in as an offset every time. Now, a couple of test loops here. 859.2.py. That's just a test loop. You use it with the terminal. On the test711 is another test loop. These are all available and downloadable. Okay, we upload to thingspeak every six minutes. Thingspeak's free service for low usage. Takes data, presents it on a series of charts, on an HTML page, so you can look at it on your browser, anywhere in the world, on your mobile phone, anywhere. You can average the inputs to give different presentations. So the top left that we looked at was the days' worth of weight, but below that, there was 10 days' worth of weight. Same information, just presented differently, gives you different information. So let's look at an interesting day in July. Okay. Right, well, this is in the heat wave. Bottom right, you can see the temperature bombing up to 30 yard each day. And if you look at the lower left, you've got a nice weight buildup. It's building up from 60 to over 80 kilograms over 10 days. That's a lot of weight. At its peak, they were pulling in about four kilograms a day. That is a lot of honey for one beehive. Top left is interesting. You can see 0800 on the left side of the x-axis. They go out about that time, well, they were out earlier than that in the heat wave, but they don't bring much back till about four o'clock. I suspect they didn't go out very much because it was a bit hot. And then in the evening, they pulled everything back. They didn't pull anything back in the day. It all came in in the evening after four o'clock. And then they pulled back. What have we got here? Yeah, about over two kilos. That's not bad for an evening shift. So interesting. It was so hot during the day, they didn't bother. Top right is interesting. Well, I think it's interesting. I don't know whether you can see the laser pointer here, but this is the brew temperature. They're maintaining it between about 33 and 34 degrees. So when it zooms up because the temperature outside is high, they will cool it down by fanning and they'll bring it down. It'll sit there and then it'll go up again. So they've got to keep up and down, but they maintain it within half a degree of optimum, which tells you there's a one-degree dead band in the bee's temperature measurement system, which is just as well otherwise to be working their socks off, trying to achieve control that they don't need. One-degree is fine. The part I don't understand about this curve is that the hysteresis reduces as the heat wave progresses. I keep thinking about that one, one they'll understand it, but it shows how much you can get out of these charts. You can spend quite a lot of time studying them and thinking about them, as I do. Okay. So this is about the temperature charts we've looked at. I think we've covered all this. Yeah, okay. The hive was getting full. Two problems there. It could induce swarming. And beekeepers do not like swarming. All your workers disappear. Half your workers go away with the queen. You've lost your honey gathering potential. So beekeepers do lots of cunning things to try and get over that problem and try and preempt it really. So that curve of the gradual increase in temperature also tells me it's time to spin the honey, time to pull off the supers, put in the centrifuge and pull the honey off. I've done very well on honey this year. Each hive has given me over a hundred jars. Right, I need to say something about an alternative way of presenting data. If you go on the internet and check out Highstock, the Highstock uses, it's a JavaScript file that resides within your browser page. And it uses the thingspeak data and presents it on a much longer time base. You can show a whole year's worth of data. So you can show historical data. This is a trace from actually the Arduino one because it's got more temperatures, it's got more sensors, it's got humidity on this. I tended to, I decided to abandon that. It wasn't doing me any good. But when I went to Raspberry Pi, I forgot the humidity. So this is one day's worth, and on the bottom here, you can drag that and you can have a variable time. You can pull down more data. You can have a week's worth of data, a month's of data. You can have a whole year's worth of data if you've got the patience to do with it. And just to show that it's possible, this is one year for my two hives combined using PaintShop Pro. So it's two traces. Very similar sort of events are going on, although each hive is different. And there's a lot of detail there that won't show like when they swarmed when I put supers on, took the supers off. But generally, we're coming out of winter, it's building up, that's stepped there. I put some supers on, another super going on, and then a really good honey flow coming here. Took loads of honey off here, and then it goes a pizzas away, and we're back into August and autumn and heading into winter. So the bees are always planning ahead. But that's the information you can get off. Hi, Charles, really interesting stuff. Okay, I was gonna do a plug for PaintShop Pro. Forget that, we're running out of time. This is a classic swarm trace. Okay, there's several pointers. Always happens May or April, sometimes June, July, but I've never experienced that. There's a peak April, May. And usually about midday, and you can see on that chart there's a one kilogram drop. And that one kilogram isn't the waves of the bees, that's the fuel. They're all tanked up on honey. They're away with the queen. Half the colony's gone off with the queen, and they were sitting on a tree or a fence, very local to me when that happens. I've got the parameter that's always the case, they always do it when I'm all on it. I don't know how they manage that, but they're pretty good. Okay, another trace. This is high speak coming out of winter. The danger time in winter is about March because they're eating food, they're preparing the brood, they're always planning ahead, they're getting thousands of young bees ready for the spring. So when the pollen's around and the nexus around, they've got the workforce available. But they need to get that workforce ready in the winter. And of course that means they need to get temperature up to 34 degrees, and that means they need to consume a lot of food. So the danger time is around about here. Coming out of winter, April I think that is, I can't read it backwards. And then magically they take off. There's food coming in. But of course with this system, if I see that weight is getting a bit low, I can give them some food. I can give them some hard food or I can give them a liquid feed. Right, so that's about it. Future work, possibly image recognition. So you can detect swarm patterns on the fronts of the hive because there are certain things. You know, the front of the hive will be plastic, it'll be with bees, you can do image recognition and you get advanced notification. With ThinkSpeak, I think you can send an alarm to your mobile even. So you could do audio analysis, microphone for airborne sound, you have an accelerometer mounted on the wood of the frame for frame-borne sound. I'll finish off with one sound clip, which may not come up. There should be a hyperlink on that. Hello, there we go. Let's have a big round of applause for Joe while I get this up. For the last screener. I want the last screen because of that. Oh, you want the last screener. Okay, yes, unfortunately we're over time so there's no time for questions. I'm sure Joe would be happy to answer your questions afterwards outside the tent. So thank you very much and really appreciate your time and attention. Please remember we need volunteers, so please volunteer at the InfoTent or the Green Room if you have any time at all. I have been volunteering today being the Herald on stage today and I would really appreciate it if any of you who have the time would volunteer as well. Thank you very much.