[elmer_fud]

Hi Everyone,

I have had a few questions about the data logger that I have on my hives, the design of it, and some of the data. I am starting a thread covering my system, the data, and what I would do different. I am going to initially grab the first several posts as place holders, one post per data set/function. I am doing this because I think there is a 4/5 picture per post limit.

Keep in mind that all of this data is from a single hive at a single location. I started beekeeping in 2017, and started the data logger running (about a week) before I got my first package. I have learned what I am doing a lot better since I started and some of this may show up in some of the things I have done. Having data on a hive does not replace a proper inspection, and will not tell you if a hive is crashing due to mites, is queen less, or has some other problem(s) going on.

If you have any requests for specific data sets, generic questions, or anything else ask away, and I will see what I can do.

The data graphs are mostly in 2 formats, one where I display all of the data sequentially, and one where I display it based on the calendar date and overlay each year. In addition to the long term graphs I also have graphs for most months, but I have found the long term (year over year) data more interesting in most cases.

Unfortunately I lost this hive in October 2019, so the data sort of flat lines/follows ambient at that point. I lost this hives due to mites, but that is a discussion that I have covered elsewhere and do not want to go into detail about in this thread.

Enjoy, and hopefully this is insightful and interesting to some of the people on here.

 

[elmer_fud]

part 1 - Hardware

I decided to build a home made data logger because it seemed interesting and I had seen someone else make one. The one that I saw that I saw and built upon (added additional features) is documented here: https://makezine.com/projects/bees-sensors-monitor-hive-health/

The sensors that I have in my data logger system are:
1. A scale under the hive
2. A temperature sensor embedded in a the middle of a frame
3. A temperature sensor on the inside of the inner cover/quilt box on the top of the hive
4. A humidity sensor on the inside of the inner cover/quilt box on the top of the hive
5. Several temperature sensors outside the hive to measure ambient/outside temperatures
6. A humidity sensor outside the hive to measure ambient/outside humidity

I have the following hardware within the system to measure/record/interpret the data from the sensor:
1. A data logger board - This board has a low resolution (8 or 10 bit) A2D, a SD card to save data 2, and a few digital communication ports.
2. A battery backed up real time clock board - This is used to keep track of the time and to generate the time stamp for the data. Since this board is battery backed up it should not be affected by power cycles that usually cause clocks to reset
3. A breakout board to measure the data from the scale. This board has a high resolution (20 or 22 bit) A2D adapter to convert scale readings to a digital values
4. A custom PCB for power supplies, voltage divider circuits, and other features as needed
5. A logic level converter

This is what the electronics in the system look like:

This is the frame with a temperature sensor embedded in the middle of it.

This is what the temperature sensor and humidity sensor on the bottom of the inside cover look like. The metal probe is the temperature sensor. The humidity sensor is in the pocket on the side behind the metal screen (and teflon tape, that did not work).

 

[elmer_fud]

Part 2 - Weight measurements

Under the hive I have a scale. The scale is a postal scale rated for 400 pounds and wired to the scale breakout board. Since the scale is under the hive it is measuring the weight of everything in the stack including the bottom board, all the brood and super boxes, the inner cover or quilt box, the telescoping lid, and the cinder block on top.

I believe the scale is still reading within about 5% of the actual value because it shows a change of about 20 pounds (+- 1 to 2 pounds) when I remove the 20 pound cinder block from the top of the hive. When I have completely removed everything from the hive the scale reads near zero (within 5 pounds).

This graph below is of the weight of the hive over time. Each color is a different year. Most of the step changes were when I added/removed equipment, or there was snow sitting on top of the hive. The single low data points are usually where I was doing inspections, so you want to look at the solid continuous lines to get a representation of the hive weight.

During the summer I can see daily cycles of the bees leaving and returning to the hive. During bad weather and the winter I do not see the daily weight cycles. The lack of weight cycling during the winter also makes me think that the scale I am using is not severely effected by temperature.

This is a graph of the weight with a bunch of annotations of what happened in/to the hive. It is from a presentation I did last spring so it does not have as much data, but there are some interesting correlations.

It is also interesting to note that the winter stores consumption is not constant, it can vary a lot based on hive strength, weather, and buildup. I do not have data to show this but during the 2019/2020 winter my stronger hive (that made it) went thru probably 50+ lbs of stores, and the weaker one went thru less than 30.

 

[elmer_fud]

Part 3 - Temperature Measurements

I am recording/logging the temperature in multiple places. I am using thermistors to measure the temperatures.

I have a temperature sensor embedded in the middle of a frame in the bottom deep that is used as a brood box. This temperature sensor has had problems. I replaced it once, but it went wonky again. I am not sure if something in the hive is messing with it, if the cable is just getting damaged, or what is going on. Take the data from this sensor with a grain of salt since it has gone wonky twice, but is still shows some interesting trends when it works.

Last spring the temperature sensor embedded in the middle of a frame looked like this. This picture was in April so the bees and brood chamber had moved up in the hive out of the bottom deep.

This is what the temperature in the middle of a frame looks like over time. It looks like the temperature runs near 95F when there is brood in the frame, and it drops during the winter when the brood nest moves up in the hive and off of the temperature sensor. The below zero (or other cold) values during the summer are when the temperature sensor was being wonky, so take the data with a grain of salt.

This is what the temperature at the top of the hive on the bottom of the inner cover/quilt box looks like. This sensor has not gone wonky, but it sees more seasonal variation since it may be farther from the brood nest at times.

This is what the outside/ambient temperatures looked like. I don't think the temperature actually got above about 110 outside, but I think the sensor may have picked up some extra heating during the summer sitting in the sun surrounded by fences.

 

[elmer_fud]

Part 4 - Humidity Measurements

There are 2 humidity sensors in my system. One is located outside the hive to monitor ambient humidity, and the second one is located on the bottom of the inner cover/quilt box on the top of the hive. I think these sensors max out at about 90% RH, so when the measurements are at this value, they actual value may be higher.

This is the humidity inside the hive. I added a quilt box in the spring of 2018 after having a lot of problems with mold. After adding the quilt box my mold problems went away and the humidity in the hive was no longer near saturation. I now run quilt boxes year round.

This is what the outside (ambient) humidity looks like.

This is what the mold inside the hive looked like before I started running quilt boxes. The wood is black because I had burned the mold off with a torch at some point in the past.

 

[elmer_fud]

Part 5 - Oddities/other findings

In some of the data it looks like the brood actually cools down occasionally and then warms back up. I am wondering if the larve cool down a bit from water evaporation when they are being fed, if they cool down at the end of the development cycle, if I am picking up wind blowing into the hive, or if my sensor is just being wonky on occasion. In the graph below you can see the outside (scale board) temperature has daily cycles, but occasionally the temperature in the middle of the frame also drops then comes back up.

During the winter I am not seeing the daily weight cycles that I see during the summer. This is what you expect when the bees are all clustered in the hive and not flying. I think there was snow on the hive from the 22 to the 30 of December in the graph below:

 

[elmer_fud]

Part 6 - Changes if/when I make another data logger and any other thoughts

Based on having this data logger running for several years and things I found building it I would make the following changes if/when I make another one in the future:
1. Move all of the electronics to one board and clean everything up
2. Use a different temperature sensor in the middle of the frames since the ones I chose are not surviving very well
3. Log the data at a much lower frequency. Right now I am logging the data at about 4 times per minute, and this is way to fast. I have many gigs of data and it takes a while to manipulate it into usable data sets or graphs. I also don't think the hive changes rapidly very often other than during inspections and maybe swarms (which I have not had yet)
4. Maybe add a oxygen and/or CO2 sensor. I have seen someone mention that they were looking at this for hive health during the winter, but I am not sure if anything ever came of it. This might be an interesting additional parameter.
5. Maybe make the system solar powered. Right now I have a 12VDC line running from inside my house to the data logger. It draws about 100 ma, but I suspect I could get this value down if I wanted to.
6. Assorted code changes to make it easier to download the data off the system. The procedure is a bit wonky right now but it works. I got the logger running about a week before getting a package, so I stopped messing with it at that time.

 

[elmer_fud]

Part 7 - Reserved for future additions #1

 

[elmer_fud]

Part 7 - Reserved for future additions #2

 

[Hogback Honey]

Wow, that is a LOT of info, and a lot of work. Appreciated though, very interesting.

 

[jonsl]

Nice project. Just curious, where did you get the frame shaped plastic container to house the electronics?

 

[Litsinger]

Wow, that is a LOT of info, and a lot of work. Appreciated though, very interesting.

Agreed, elmer_fud. I have appreciated your insights on this type of technology.

Do you have immediate plans to incorporate some/all of your lessons learned into a v.2 system? I noticed post #7 said (if/when).

 

[Trin]

That's a lot of work! Appreciate you are doing it.

Maybe interesting to measure daily sunlight hours. It seems that some days the bees mostly stay inside further into the day until the sun actually shines. So I'm wondering how cloud density/level of light effects foraging.

There is a level of complexity with sunlight. Do bees require it for chemistry/metabolism reasons? (Thinking about how sunlight influences vitamin D levels in humans).

Wind speed and direction and duration? If I had a weather station I would donate it.

All the same there might be some correlations respective to other data points you are collecting.

How many lifetimes do you have?

 

[elmer_fud]

Nice project. Just curious, where did you get the frame shaped plastic container to house the electronics?

I got the box for the electronics from Digikey.

This is the box I am using. If you search thru the boxes/enclosure section on digikey there are lots (1000's) of options.
https://www.digikey.com/products/en?keywords=377-1764-ND

 

[elmer_fud]

Do you have immediate plans to incorporate some/all of your lessons learned into a v.2 system? I noticed post #7 said (if/when).

I have designed a version 2 that has all of the components on one board, but it has been atleast a year since I did anything with it. I never built up version 2, so it is a long ways from being functional. During the spring/summer/fall when it is nice outside I would rather do stuff outside than sit in the basement on my computer. I have been working on some other projects, but maybe I will get back to V2 next winter.

 

[elmer_fud]

Maybe interesting to measure daily sunlight hours. It seems that some days the bees mostly stay inside further into the day until the sun actually shines. So I'm wondering how cloud density/level of light effects foraging.

There is a level of complexity with sunlight. Do bees require it for chemistry/metabolism reasons? (Thinking about how sunlight influences vitamin D levels in humans).

That is an interesting thought. I may have to look into that if/when I ever to version 2.

Wind speed and direction and duration? If I had a weather station I would donate it.

All the same there might be some correlations respective to other data points you are collecting.

I suspect my hives do not get much wind. There is a 6 ft fence on 2 sides, a 5 ft side of a sunporch on one side, and a 20+ ft pine tree on the fourth side. I could put a weather station on a pole, but it would have to be a ways away from the hive or 20+ feet up to get any reasonable data.

This is where my hives are located (with a tree at my back)

It is not a great spot for winter sun, but it is good for keeping the bees and people separated. Between my house and my neighbors is all rock and there is not much traffic in the area.

How many lifetimes do you have?

I only have 1. I don't have any kids (yet) which means I have a more free time. I developed an interest in robotics in college and have built several robots in/after college. I am a mechanical engineer by degree, but I do a bit of everything now.

 

[Litsinger]

During the spring/summer/fall when it is nice outside I would rather do stuff outside than sit in the basement on my computer. I have been working on some other projects, but maybe I will get back to V2 next winter.

This I do understand, and I am impressed that you were able to take-on such an ambitious and well-thought-out project. Hoping to read about your updated monitoring system maybe next Spring. I'll stay tuned...

 

[jonsl]

I got the box for the electronics from Digikey.

This is the box I am using. If you search thru the boxes/enclosure section on digikey there are lots (1000's) of options.
https://www.digikey.com/products/en?keywords=377-1764-ND

Thanks for the info! I may do something similar now that I have a lot more time on my hands.

 

[jcahn2]

Fascinating project Elmer and the data are quite valuable. This can only be more useful as time goes by and more data recorded. Thank you.

 

[elmer_fud]

This can only be more useful as time goes by and more data recorded.

I agree.

A coworker sent me a paper from a college that made a beehive data logger. They talked a lot about the system setup, and how they designed the system for a cloud interface, and then had 2 months of data during a flow at the end. The 2 months of data did not really show anything useful or interesting.