Feasability question beehive data logger

Hello all,

I'm a beekeeper wanting to measure hive conditions (starting with basic temperature, humidity, and battery level). I know a lot of people have done this, and there are even numerous commercial products out there to achieve this, but I want to do it myself. Something like this would be what I'm going for, although smaller would be preferable:

Minimum requirements:

    1. Battery powered and will work for at least 6 months.
    1. Polling the sensors at least twice per hour, maximum of four times per hour.
    1. Transmit data through WIFI (doesn't have to be every hour, once a day minimum, ideally hourly).
    1. No cables in the hive.
    1. Data ends up in a MySQL DB.
    1. Accurate timestamps.

I've done a bit of research and to me there seem to be 2 plausible options:

  1. Have a simple datalogger with RF module that sends data to a HUB, and have the HUB deal with internet. This makes the battery lifetime relatively easy, and would make it cheaper to scale up if I want to monitor other hives at the same time, or add other sensors (like a hive scale or bee counter or sound frequency). The HUB could be an ESP8266 or even a raspberry pi.

  2. Make it as a stand-alone device, which I like for the 'elegance' and simplicity of the end product, but seems a little less practical in scaling up and harder to produce / program. My biggest issue with this option is maximizing battery lifetime while having to power a WIFI-module. That said, I'd still like to explore the options of this method.

The way I could see this work right now is: some basic barebones atmel chip as master that periodically gets woken up from deep sleep by an external RTC, collects data, and then periodically powers up ESP through mosfet and sends the data over WIFI. I'm not sure how realistic it is to run something like this for 6 months on batteries though.

I'd love to hear suggestions / ideas / recommendations / criticism to make this project a reality. It's gonna be challenging and I have plenty to learn, but I feel I can do it once I decide which route to take.

Best regards and thanks for any input,

Ramon

Now that thing has a nice price tag on it. I'd go for the esp8266 and have arduino run on it. Sensors added via i2c or spi. 15 minitues into deep sleep, then wake up .. this might help: https://openhomeautomation.net/esp8266-battery

Last time this question was asked - a few weeks back - I suggested that the bees might go and plug up the holes on the sensors. :astonished:

Thanks guys for the feedback, sorry for a late response.

@zwieblum, At that price I wonder how many units they sell. Beekeepers generally speaking are notoriously thrifty. Thanks for the link as well.

@Paul__B, I asked that question to one of the companies that make the commercial product and they say it doesn't matter too much. The sensors will still work as long as the wood picked for encasing can 'breathe'.

I've decided to use ESP-NOW and create a gateway (ESP-32), and then have the ESP-8266 as a controller. That way I save precious time and battery life on the ESP-8266 as it's much faster to send the data that way than having it connect to WIFI.

For the RTC I've chosen the DS3231. From what I've read I can use a pulse to the reset pin of the ESP-8266 to get it out of sleepmode.

I know how to do most of the parts it's just about bringing it all together! :slight_smile:

IHAVEBEES:
2. Make it as a stand-alone device, which I like for the 'elegance' and simplicity of the end product, but seems a little less practical in scaling up and harder to produce / program. My biggest issue with this option is maximizing battery lifetime while having to power a WIFI-module. That said, I'd still like to explore the options of this method.

One of the issues here would be that the 'product' is restricted to being withing WiFi range, which may not be very far.

The other issue is that WiFi setups are not very battery friendly, relativly large amounts of power being required.

Then there is LoRa, which if used in the 'product' would allow for hives to be widly spaced since very low powers can cover very large distances. The nodes in the hives could have battey lifes of several years.

Yes you would need a central hub of some sort that is connected to the Internet to recieve the data.

BTW: What would be an acceptable price tag on a unit like that for a beekeeper?

srnet:
One of the issues here would be that the 'product' is restricted to being withing WiFi range, which may not be very far.

The other issue is that WiFi setups are not very battery friendly, relativly large amounts of power being required.

Then there is LoRa, which if used in the 'product' would allow for hives to be widly spaced since very low powers can cover very large distances. The nodes in the hives could have battey lifes of several years.

Yes you would need a central hub of some sort that is connected to the Internet to recieve the data.

Agreed that this approach is better. Instead of LORA I'm going to try ESP-NOW first though. This is not for a commercial product but for myself. My hives are close enough to a WIFI network. If I can make it work with ESP-NOW, I'm assuming that doing it with LoRa later if needed isn't that much of a hassle. I also like the modularity of this approach more.

zwieblum:
BTW: What would be an acceptable price tag on a unit like that for a beekeeper?

Hard to give an exact number, but I know that a lot of beekeepers are up there in age and do things old school, and the younger ones seem like the DIY types.

Speaking for myself, 50-60 euros MAX for a stand-alone WiFi device.
For the hub and controller setup, maybe 60 for the hub and around 40 per controller?

I know it's near impossible to produce for that low with all the development costs, but still, 189$ like the beebot is way too steep IMO.

A scale that measures the weight of the hive and publishes that to internet would be more interesting to beekeepers since it tells you more about colony health than just temperature and humidity. I've seen those retail at 350$ and up though. I guess if you can make something like that but a bit cheaper that would be pretty interesting for lots of beekeepers.

Interesting. Is that for a finished product or a kit?

What do you get from weight? amount of produced honey? And what do you get from temperature/himidity? (Health of bees?) (Sorry - I never went that deep into beekeeping)

I used to have the same mindset on how expensive some consumer electronics can be. I’ve built my own garden management/monitoring solution soup to nuts (LoRa to www).

I can attest that $200 for the BuzzBox is very reasonable. In consumer electronics, unless you can 1) Sell thousands at low margins 2) Sell fewer at high margins, you’re going to be a “hobby”.

Not trying to discourage you, but what appears to be expensive with niche products is actually not. It’s unfortunate that agriculture in general is a very appealing field to add tech, but as you mentioned most farmers/beekeepers are very cheap and would rather do things the way they’ve been done for centuries :confused:

zwieblum:
Interesting. Is that for a finished product or a kit?

What do you get from weight? amount of produced honey? And what do you get from temperature/himidity? (Health of bees?) (Sorry - I never went that deep into beekeeping)

That's for a finished product. Some you need to get additional subscription as well to use their webserver.

Temperature / humidity: Bees are very apt at temperature regulation. They keep their broodnest between 32 and 35 degrees celcius, regardless of outdoor conditions. So a healthy hive will show something within this range. If there's no queen, after a while they'll become broodless, and temperature starts to fluctuate more. Then if you introduce a new queen for example, you will know she's laying when the brood temperature stabilises again, without having to disturb them. See graph from some other beekeeper below:

Beehive weight: Can tell you whether a colony is growing. They are at their smallest right after winter, and then start growing until roughly july. You can assess productivity (honey collection), sudden drop in weight could mean they're being robbed or swarmed. Could help identify when feeding is required etc.

Lots of uses :-)!

czu001:
I used to have the same mindset on how expensive some consumer electronics can be. I’ve built my own garden management/monitoring solution soup to nuts (LoRa to www).

I can attest that $200 for the BuzzBox is very reasonable. In consumer electronics, unless you can 1) Sell thousands at low margins 2) Sell fewer at high margins, you’re going to be a “hobby”.

Not trying to discourage you, but what appears to be expensive with niche products is actually not. It’s unfortunate that agriculture in general is a very appealing field to add tech, but as you mentioned most farmers/beekeepers are very cheap and would rather do things the way they’ve been done for centuries :confused:

I agree that from a production point of view it's not too expensive. From a beekeeping perspective (making money) it is too expensive for the added value it brings to your honey production.

I do see value in research for these things though.

For me, the reason to make it myself is because I want to learn more about electronics and it's fun to do. I also work in science, so measuring stuff is right up my alley!

Hm. From the science-side I'm absolutely with you. But from the economical side, what can you gain from this knowledge?

zwieblum:
Hm. From the science-side I'm absolutely with you. But from the economical side, what can you gain from this knowledge?

Well, since the costs are too high I don't think any commercial beekeeper will opt for this. That said, there have been a lot of European grants gifted to projects within this area. Like millions of euros.

If devices were much cheaper, you could probably spend less time managing your hives. Another thing is theft / vandalism. Many beekeepers would love to know when their hives are being moved or have been knocked over. So some of these devices contain GPS with cellphone plans as well.

IHAVEBEES:
If devices were much cheaper, you could probably spend less time managing your hives.

Perhaps it's an opportunity then, is it possible to design, get approval, manufacture and sell a device that is 'much cheaper' ?

It’s really not going to be about the cost, it’s about the money it saves you. If it saves a hive, how much would that save?

Anyways, a passionate hobby can easily lead to a great product, but I wouldn’t suggest going into it thinking something is too expensive.

I think what you’re doing is pretty cool and I’d love to hear updates. My garden project is www.greenclaytexas.com

Is it expensive to produce? Absolutely... but I have some hella produce :slight_smile:

czu001:
It’s really not going to be about the cost, it’s about the money it saves you. If it saves a hive, how much would that save?

Anyways, a passionate hobby can easily lead to a great product, but I wouldn’t suggest going into it thinking something is too expensive.

I think what you’re doing is pretty cool and I’d love to hear updates. My garden project is www.greenclaytexas.com

Is it expensive to produce? Absolutely... but I have some hella produce :slight_smile:

That's a cool project you've got going on! Hobby that turned into business?
Thanks for the encouraging words as well! Will try to post some updates.

srnet:
Perhaps it's an opportunity then, is it possible to design, get approval, manufacture and sell a device that is 'much cheaper' ?

A good question to consider. For now, I'll just have fun and focus on trying to build it! :smiley:

Back to the project. Have data gathering / pushing it to the internet down. Realised that minute accurate timestamps are good enough, and that timestamp can drift over time. It doesn't necessarily have to be xx:00, xx:15, xx:30, xx:45 exactly.

Now have 3 possible design options:

  1. Kevin Darrah's TrigBoard method:

He uses a TPL5111 to switch a LDO based on timing interval. This approach should yield the lowest sleep current. Downsides are that you need a LDO and the gate. Timing has 99% accuracy on average.

  1. Using the DS3231

Sleepcurrent of ESP should be roughly 5-10 na. Then the RTC in battery mode should use roughly 3 na. --> Very accurate. Advantage to also add other interrupts later easily on the other interrupt pins.

  1. Using the ESPs internal timer:
    Fewer components but not very accurate at all.

Progress made: logging battery voltage (voltage divider), temperature and humidity and pushing it to the internet has been figured out (get requests to php script).

TPL all the way. Either 5111 for the LDO or 5110 for battery. The 5110 doesn’t play well with switching regulators (some chips use internal regulators).

I use the 5110 on my project but I had to add a capacitor to smooth out startup spikes. 5111 I’ve never had issue.