Contributor Covenant GPLv3 license Build Download Doc

Beelance

Overview

Beelance comes from the French word “Balance” which means “weight scale”.

Beelance is an autonomous and remotely connected weight scale for beehives 🐝. It can monitor the weight and temperature of beehives and send the data to a remote server or a Google Sheet (through IFTTT or alternative). Beelance supports 2 types of communication:

Device interface:

Example of Google Sheet receiving the data:

Features

Powering

Beelance can be powered by USB-C, USB-C Solar Panel or battery. On the field, it is recommended to use a Solar Panel to power the device. If a battery is used, the device will automatically switch to the battery when the solar panel is not providing enough power. Otherwise, the battery will be charged by the solar panel at the same time the solar panel powers the device.

Battery type: 3.7V 18650 LFP or Li-ion

Communication

Supported frequencies: CAT-M (LTE-M), NB-IOT, LTE, 4G

Normal cellular SIM cards or specialized M2M (IoT) SIM cards can be used.

Measurements

The device takes measurements and sends them periodically:

It will also keep a local history of the 10 last measurements, the maximum values of the last 10 hours and the maximum values of the last 10 days, both for weight and temperature.

Configuration

The device has an access point, we can connect to it to:

Integrations

The device can be configured to send its data to a remove server of your choice. It can also be an IFTTT webhook, which can then insert the data in a Google Sheet.

Downloads

Firmware downloads are available in the Releases page of the project.

Firmware files are named as follow:

Where:

How to build

Connectivity provider

Beelance works with a SIM card, so you need to select your carrier:

Here is a list of LTE-M / NB-IOT providers for France amd Europe:

Here is a list of cheap standard mobile carrier providers for France amd Europe:

Beelance DATA consumption

The JSON payload sent from Beelance is more or less 250 bytes.

{
  "ts": 1731147105,
  "bh": "beelance-73fadc",
  "temp": 24.56,
  "wt": 0,
  "lat": 0,
  "long": 0,
  "alt": 0,
  "sim": "8944501905220523406f",
  "op": "Orange F",
  "dev": "73FADC",
  "boot": 36,
  "ver": "main_9203b08_modified",
  "up": 103,
  "pow": "ext",
  "bat": 0,
  "volt": 4.23,
  "eco": false
}

You also need to add the HTTP/HTTPS headers overhead, which can easily be at least 100 bytes.

content-length	238
content-type	application/json
user-agent	Arduino/2.2.0
host	webhook.site

HTTPS will also add more overhead because of the handshake and encryption. So it is recommended to use HTTP.

Beelance will send the data every 1 hour by default, and only between 5:00 AM and 11:00 PM. All these settings are configurable.

With the default settings on HTTP, Beelance will consume about 6300 bytes / day, so less than 200 Kb / month, so less than 3 Mb / year.

Shopping list

Hardware

WARNINGS:

  1. Make sure to verify what you receive. The seller SAMIORE Store is quite reliable: I often buy electronic stuff at this store. I already received broken items or unfinished items from some other sellers.

  2. Verify the Solar Panel Voc (if you buy another brand): it should be within 4.4-6V.

  3. We recommend to use the T-SIM7080G if using both a battery with solar panel because the LILYGO T-A7670G does not have a good PMU and won’t switch to the battery when solar panel is producing just a little power on cloudy days or start/end of day.

Wiring

Default GPIO Pinout

  T-SIM7080G-S3 Pin T-A7670G R2 Pin
Temperature Sensor DATA 12 15
Temperature Sensor VCC 3.3V 3.3V
Temperature Sensor GND GND GND
HX711 VCC 3.3V 3.3V
HX711 SCK (CLOCK) 13 13
HX711 DT (DATA) 14 14
HX711 GND GND GND

How to wire weight cells

Assembly

The box and solar panel:

The weight sensors with the 3D printed enclosures:

Calibration

How to install

SIM card

Do not forget to install the SIM card before powering.

Powering

WARNING / DISCLAIMER:

Leaving a device in nature with a battery is a potential hazard, especially with Li-ion batteries which are not supposed to be used above 60 degrees. Please make sure you are informed with the associated risks with Li-ion or LFP batteries. If you are not sure about the safety of the device, do not leave it unattended or without any monitoring, or do not install any battery. In any case, use an appropriate box which is fire retardant (UL94V-0) and waterproof. The project’s authors decline any responsibility regarding the use and assembly of this project. You are entirely responsible of any loss, damage or injury caused by the use of this project.

Powering with the internal 3.7V battery

This option is compatible with the solar panels.

The board accepts an internal 18650 3.7V battery, but you could pick and Li-ion or LFP battery of 3.7V nominal voltage but with higher capacity. This battery will be recharged when the USB-C will be powered (by solar panel or something else).

it is also possible to use a bigger 3.7 battery and connect the VCC and GND to the internal battery holder.

To start the board from the internal battery:

  1. Set the board switch to ON
  2. On the T-SIM7080G-S3, to start from battery, you need to press 3 seconds on the power button

Powering with USB-C power bank

Another option is to use a big power bank of 20000mAh or more and use it to power the device with usb-c. If you are using the sleep option between each push, and send each 2 hours, and use the night mode to sleep the device, a power bank can allow you to last a very long time. The duration depends on the device and push frequency and sleep parameters, but it should last several weeks up to several months.

Powering with USB-C Solar Panel

This option is compatible with the internal 3.7V battery: the solar panel will power the device and charge the battery with the sun and the battery will take over as needed.

DO NOT USE the board internal solar connector. We use instead a USB-C Solar Panel which is capable of automatically power the board in case it shuts down. Never power the device with both the on-board solar connector and a usb-c solar panel.

This option is not working very well with the LILYGO T-A7670G because this board does not have a Power Management Unit (PMU) like the T-SIM7080G-S3. So when the solar panel is not able to produce enough current, the device will simply not start, until the solar panel voltage becomes 0 and then the device will switch on the battery.

Beelance consumption

Here is the power consumption of the device in 2 modes:

  1. Active mode (above / left): from when the device starts or wakes up until when it sends the data and goes to deep sleep
  2. Deep sleep mode (below / right): when the device is in deep sleep

The device consumption in active mode is on average less than 200mA. So a charged battery having a capacity of more than 3000 mAh (3200-3600 mAh) should last about 15 hours if kept in active mode

The device consumption in deep sleep mode is less than 500 uA. So a charged battery having a capacity of more than 3000 mAh (3200-3600 mAh) should last about 8 months.

By default, the device will wake up every hour between 5 am to 11 pm, send the data and go back to sleep. We will suppose that the device takes 3 minutes to register on the network. get a GPS point and send the data. (which, in reality, is much less).

The device will be in active mode for 3 minutes 16 times a day, which is 48 minutes per day. The device will be in deep sleep for 23 hours and 12 minutes per day.

So the consumption of the device per day should be no more than 220 mAh, which means about 2 weeks of autonomy with a 3000mAh battery.

All these numbers are quite theoretical and the real consumption will be different because the heat and cold will affect the battery capacity.

But coupled with a solar panel, the device should be able to last indefinitely.

Note: the testing was done with the temperature probe attached.

Firmware flash (first time)

First time, flash the entire firmware which includes the partition table and all partitions. Pick the right firmware for the right board.

  1. Erase the flash
esptool.py erase_flash
  1. Flash the firmware
esptool.py write_flash 0x0 Beelance-VERSION-CHIP.FACTORY.bin

On Windows, you can use the official ESP32 Flash Download Tool. Pay attention to first erase and then flash the entire memory by setting the flash address to 0.

Device Configuration

In the main dashboard, click to restart the device and wait to makes sure it becomes ready (modem ready, time and GPS fix, etc.). You can also go to the console at http://192.168.4.1/console to see the logs, and if you really want to see verbose logging, you can activate debug logs in he console.

Once your testing is finished and when installed, you can deactivate the option Prevent Sleep: when disabled, the device will be allowed to go into deep sleep after it has sent the data. If the device is in deep sleep and you need to access it, just power it and quickly connect and toggle the button to prevent it to go t o sleep again.

Important information about the Modem

Connecting the first time with a new SIM card can take a very long time: first 30 seconds, the app tries to automatically register the SIM with an operator. Cellular communication goes through several steps:

  1. Modem startup: can take up to 20 seconds
  2. Network registration: can take up to 30 seconds
  3. Network search: if step 2 fails can take up to 1 minute
  4. Try network registration for each search result until one succeeds (can take up to 30 seconds per operator tried)
  5. GPS search: can take up to 1 minute (timeout can be configured in the configuration page with gps_timeout)

The device includes a watchdog for all these steps: it will automatically restart if it becomes locked in a state for 5 minutes.

If the device becomes disconnected (loss of Modem IP address), it will try to reconnect automatically.

Weight Scale Calibration

The weight cells are not calibrated by default because each setup is different and will require different calibration values (called offset and scale).

When the device is installed and powered on, you will see some weird values for the weight, which can even be negative:

Follow the steps below to calibrate…

Temperature effect

This is important that you run the calibration at a time of day when you want to see more accurate results.

The temperature has an effect on the load: for example, during the night, the load will be higher because the bees are in the beehive and there is also more humidity in the air and wood. Under a steady bright sun, the temperature is increasing and the load is heating / drying, so the weight will decrease. Here is an example of a 19.8 kg load put on the weight scale for several hours under the sun. You can see that the more the temperature is rising, the more the weight is decreasing. So you need to take this effect in consideration when looking at the statistics. You will still have the same effect whether you include the beehive in your calibration tare or not.

Also, the temperature rising has an opposite effect on the metal of the weight cells themselves: the more the temperature is rising, the more the metal is expanding and the weight is increasing. But this effect is negligible compared to the effect of the bees, honey and humidity.

Offset Calibration

The first step is to calibrate your zero weight (which is called tare). You can decide to do it with or without the beehive on the scale, depending if you want the reported weight to include the weight of the beehive or not.. Press the Tare button to do so.

Once pressed, the weight should become close to zero. You can press the Tare button again to find a more accurate zero weight, but note that there will still be some inaccuracy of about 100 grams. If you end up with a weight which is changing between -50 and 50 grams, this is pretty good.

Scale Calibration

Once the baseline is set, the second step is to calibrate the scale. This ensures to get the right weight in grams when there is a load on the scale. To do that, place on the setup (on the beehive or on the scale) a load of about 15-20 kg, like a big bag of sugar or anything else. YOU MUST KNOW THE EXACT WEIGHT OF THIS LOAD.

This is important that:

  1. You know the exact weight of the big bag of sugar.
  2. The load represents the load you will measure. For example, if you have calibrated the baseline with your beehive, you will only measure the bees and honey. So the weight will be lower compared to if you had calibrated the baseline without the beehive, and then put the beehive on the scale later.

When placing the load on the scale, the numbers will change to something strange and could even be negative. For example, I’ve put a total of 11.6 kg on the scale and here is what I see:

The next step is to move the cursor to place it to the exact value of the load you have put on the scale. For example in my case I set the cursor to 11600 g.

Once the cursor is set to the right position, after 2 seconds, the device will resume the measurements and you will see the weight changing around the calibrated value. If it becomes too high or too low, move the cursor again to the expected weight value.

Wait for 24 hours. for the meta weight cells to adapt to the load, and check the weight value again. If it has changed too much (by about 1kg for example), just set the slider again to the right value.

Under the hood

The statistics section of the dashboard will show you the calibrated offset and scale values. The tare value represents the weight that is not accounted for in the weight calculation.

The offset and scale values are saved on the device in the configuration, which can be backed up and restored and accessed in the configuration page at http://192.168.4.1/config. You can even manually adjust these values.

How to use

Dashboard

The Dashboard is located at http://192.168.4.1/. For there you can see the device status and access some basic buttons.

The statistics menu shows some metrics regarding the device itself.

Other pages are available:

Update the firmware

The firmware file to use for the OTA Update is the one ending with .OTA.bin.

It is always recommended to update the firmware to the latest version. Download the latest version on your mobile phone or computer, then connect to each device on their WiFi access point. Then go to http://192.168.4.1/ and click on the button called Update Firmware in the Web interface.

The device will automatically reboot in a SafeBoot mode (you can read more about the special SafeBoot more here).

Try to find an access point name SafeBoot-XXXXXX and connect to it: you should see the SafeBoot interface.

Go to http://192.168.4.1/ to open the SafeBoot interface allowing you to update the firmware.

Configuration reset, backup and restore

If you need to backup, restore or reset the configuration to factory settings, go to http://192.168.4.1/config and click on the corresponding button.

The same page can be used to change your configuration and some more advanced settings.

Factory reset

To do a factory reset, proceed with the Firmware flash (first time) step.

Debugging and logging

The Web Console can be opened at http://192.168.4.1/console.

To see more verbose logging including the modem AT commands, you can activate the debug_enable switch in the configuration page.

WARNING: the console allows you to interact with the Modem internals (called AT commands). DO NOT SEND ANY DATA UNLESS YOU KNOW WHAT YOU ARE DOING.

Receiving the data

JSON Payload

Here is below a sample of the JSON payload that the device will send to the configured URL in a POST request.

{
  "ts": 1712578315,
  "bh": "Ruche-01",
  "temp": 24.25,
  "wt": 15000,
  "lat": 44.1234,
  "long": -2.1234,
  "alt": 7.4,
  "sim": "89457300000014000000",
  "op": "20801",
  "dev": "73FADC",
  "boot": 1358,
  "ver": "v1.2.3",
  "up": 10,
  "pow": "bat",
  "bat": 93,
  "volt": 4.13,
  "eco": true
}

The payload is about 250 bytes.

Here are the possible combinations for pow, bat and volt:

webhook.site

https://webhook.site allows to capture the JSON payload and see it in a nice interface.

Copy your unique hook URL (example: https://webhook.site/53cc6f4b-145d-49b2-8518-7cb177b0166f) and paste it in the Beelance configuration page in the send_url.

When Beelance sends some data, you will see the full HTTP request in the webhook.site interface.

IFTTT Integration

If you have an IFTTT account, you can create an IFTTT applet to receive the JSON payload and insert it in a a new row in a Google Sheet spreadsheet. To do that, create a new applet with a webhook. The URL of the webhook has to be configured in the device configuration page in the send_url parameter (try use http instead of https).

Here is the filter code you can use to transform the JSON data into cell values. This is optional: you can also receive the JSON in Google Sheet and process it in Google Sheet.

var payload = JSON.parse(MakerWebhooks.jsonEvent.JsonPayload);
var formattedRow = "";
for (var k in payload) formattedRow += `|||${payload[k]}`;
GoogleSheets.appendToGoogleSpreadsheet.setFormattedRow(
  formattedRow.substring(3)
);

Once the applet is in place, every push from the device will reach IFTTT which will feed the Google Sheet. From there, you can create graphs and process the data as you wish.

Developer guide

Recommended IDE: Visual Studio Code with the PlatformIO extension.

Project structure

Building and uploading the firmware

pio run -t build -e <env>
pio run -t upload -e <env>
pio run -t monitor -e <env>

Contact

If you have any question related to the project, please use the project’s discussion forum. I won’t answer any project-related question by email.

To report any bug or request a feature, please use the project’s issue tracker. I won’t answer any project-related bug or feature by email.

If you want to contact me for a partnership or schedule an assembly session, please send an email to beelance@carbou.me.

If you would like to buy a pre-assembled box (optionally setup with IFTTT and Google Sheet), please send an email to beelance@carbou.me. But I strongly suggest you to build it yourself. Since I do not maintain any stock and don’t have a way to cut cost, building a box for you will be 1.5-2x more expensive than doing it yourself.

Related projects, guides and articles