1

I'm new about Raspberry world..

I'm studying a way to implement a custom beehive monitoring system.

I would like to make a premise: I'm studying to organize the project, I'm not yet implementing or testing anything.

The most important monitoring parameters are (for the moment): weight, temperature and humidity. I have my bees into a remote place so I haven't power connection and wi-fi.

To try to solve the power problem my idea is to use a battery with a solar panel and, to optimize the power and solar charge I will switch on and off the system periodically just to take the measurements. My idea is to use a similar approach like this: https://www.arrow.com/en/research-and-events/articles/solar-powered-raspberry-pi-with-attiny85

And also, just to try to reduce (as possible) the consumes my idea is to use a Raspberry Pi Zero.

About the data-connection my idea is to use the module to use a SIM (GSM/LTE) on my Raspberry.

From your point of view, are these aspects valid?

Now I have to analyze two additional big points:

As explained, I am interested in monitoring these 3 parameters so, probably, I'll need:

  • 1 temperature sensor
  • 1 humidity sensor
  • 4 weight scale sensors (one for each angle of the beehive).

The question, about this point, is: can I connect all sensors on the same Raspberry? Are the available pins enough? In general, If it's necessary to connect a lot of components, Is it possible to use the same Raspberry (is it possible to reuse a pin?) As you can see, I'm not an electronic expert (I'm a software developer) so, any help or any example/tutorial about it is vital for me.

The second aspect, similar to the first, is this: I have more beehives that are close together (I have, for the moment around 6/8 beehives). For those who don't know, a beehive is a sort of box like this: Image The beehives weight could be around 30-50kg).

I'm focused on a prototype but in the future I'll need to monitor these 3 parameters for each beehive. Obviously I need to reduce, as possible, the number of the components because, as described, I don't have the power connection and wifi data, so I would like to avoid to replicate the schema for each beehive: I would like to avoid to have multiple battery, multiple mini solar panels and multiple SIMs. Is it possible to make only one main module (only one Raspberry) with many sensors? Is it a possible real scenario or it's only into my science fiction?

About the weight scale sensors I found this project just to give you an idea. But probably I need to change the sensors because I will need to support different maximum weights.

https://tutorials-raspberrypi.com/digital-raspberry-pi-scale-weight-sensor-hx711/

4
  • 2
    Raspberry Pi is an overkill ... something like an arduino uses less power and is more than adequate for the job ... google arduino weather station
    – jsotola
    Apr 30 at 14:25
  • @Safari, I think your project plan is very good. I would like to suggest the following: (1) For prototyping, it is much more time efficient to use Rpi4B as the development computer. My estimate is that Rpi4B is 10 times more efficient that RpiZ. (2) Try the temperature and humidity sensor first, they are 10 times easier than the HX711. (3) How practical is your project plan depends on your knowledge and skills. If you only know (a) Ohm's Law, (b) how to use a multi-meter, (c) Arduino blink a LED, and not much else, it might take you at least 6 months full time to complete the project.
    – tlfong01
    Apr 30 at 14:27
  • @jsotola I agree with you, Arduino has these advantigaes… btw Raspberry, from configuration point of view it’s more easy to configure/implement… for example it’s possible to use python… using Arduino, for example, all about comunications on GSM/LT + using RestAPI (http) it’s not easy to use… I’m not expert this is my poit of view about the choise
    – Safari
    Apr 30 at 21:25
  • You can use one Rpi4B and a bunch of Picos.
    – tlfong01
    May 1 at 0:40
1

You can also combine a Raspberry Pi with an Arduino. Arduino for reading the sensors and the Raspberry Pi for processing and uploading the data. There are Arduinos with more pins than the raspberry. You can reuse the Ground Pin and I also think to some extent the power pins.

Think if you really want cables between all your beehives... You will need to make them pluggable if you ever want to move a hive. (Or does this never happen?) You also might lose some power, depending on how long the cables get. I personally just really like decentralized systems.

Even if you decide to give each beehive its own "system" you can maybe connect them via Bluetooth or Wi-Fi and designate one as the "uploader". Because those GSM Modules are expensive! :)

As to your final question: This is definitely possible depending on how much time and money you are willing to put into it. As always, I would start really small, maybe just measuring the weight, then the other data, and then step by step adding solar panel and GSM-Module and so on.

There is never one "correct" way, but I hope I could give you some Inputs!

1

I would propose to drop the Raspberry Pi and just use a Arduino.

The solution you describe isn't that challenging and Arduino can easily handle the task by itself. And Arduino has a feature "Sleep/Deep sleep" that is energy efficient since it takes a couple of seconds to wake up and a Raspberry Pi takes a much longer time to boot and start a operation.

And some sensors is analog, and a Raspberry Pi has no analog interface, and Arduino does!

And I could go on, but I think you get the idea.

As a final point, have a look at this or something similar: https://www.arduino.cc/en/Guide/ArduinoGSMShield

1

You may be making this harder than it needs to be - or maybe not. But in any case, minimizing the re-invention of the wheel, you may want to check out Cypress Semiconductor's Solar Beacons. The hardware, demo software & "other stuff" is available as a kit. Monitoring beehives sounds like an appropriate application for the Solar Beacons.

2
  • I did not know these SolarBeacons... very interesting project. If I understand correctly, I could combine these sensors about temperature and humidity + custom approach about the weight monitoring that could also represent a sort of gateway for these sensors. But I'm not totally agree with you that monitoring beehives sounds like an appropriate application for the Solar Beacons: I need to monitor the temperature/humidity inside the beehives where there isn't light.
    – Safari
    May 4 at 10:21
  • @Safari: It would have been good to have mentioned that (no light) requirement, but then I should have guessed that would be the case. In any case, I'm not sure that's a concern since most temp & humidity sensors don't need to be immediately adjacent to their power source. Good luck with this - maybe you can find out the cause of the drop in bee populations.
    – Seamus
    May 4 at 22:21
1

Interesting questions and great project! I also think that a more light weight single board computer/microcontroller would make sense for the hive sensors, but a Pi could function as an effective central hub for data processing and storage until you are able to retrieve it.

As for your questions around power/solar use: I am currently working on a sensor cluster for ant colonies, and am basing a lot of my initial design ideas off of the Open Source Beehives Project. They use a Particle Photon, more similar to a Pi zero as a small, compact IOT device with wifi capabilities running on 3.3VDC. They successfully use it with a solar panel and a bunch of sensors (without running out of pins!) and all the parts are laid out in detail if you want an example of what works well with bee hives.

As far as your concern about minimizing pin use: Like I mentioned above there are some great open source examples that get around that issue, but they also have really customized their PCB design and overall component scheme. If you are really trying to minimize, I have jammed a bunch of digital sensors and controllers relying on PWM pins onto an Arduino by making use of I2C communication bus cables, which could be worth checking out. This is especially true if your microcontrollers for the individual sensor clusters/hives won't have wireless capabilities.

If you were to try and condense as many sensors as possible to a single Pi I have to say the fact that you want 4 sensors for weight per box is what blows through your pin availability (B+ having 26 programmable pins), not even a Pi Z. You could make a multiplexer to use more sensors on the same number of pins, but if any of the sensors you want to use are analog a handful of pins are going to get used up just to make an analog-to-digital converter (assuming you go the Rpi route as they can't handle analog input directly). Weight is not a common metric I see measured, is it to monitor potential yield/wet weight? Unless it is of primary importance to you, I would say just follow your regions regular seasonal harvesting schedule/check when you go to collect the passively stored data. Without the four weight sensors you are free to add some other helpful sensor while saving pins, e.g. a wind sensor. That, along with temp and humidity, would put you at only 18-24 sensor units.

At the end of the day, to really build a light weight and efficient sensor cluster you will have to find simpler microcontrollers that can get the job done, but Raspberry Pi's can play a powerful role in coordination with them! Just an example of something REALLY cool that a Pi has the unique capability to do compared to simpler systems (and that OSBeehives implements externally): you could collect audio data from within the hives and send it a Raspberry Pi to process it and analyze it using machine learning in order to discern meaningful patterns in hive activity and detect colony health problems early!

Here are two amazing resources for further examples:

The HiveEyes Project and Compiled Project List

An aggregated list of of electronic colony monitoring system vendors.

Not the answer you're looking for? Browse other questions tagged or ask your own question.