B-Smart

Li Mo, Zuo Xiying, Hadrien Gourdet, Mohammed Merkache

"If the bee disappeared from the face of the earth, man would have only four years to live." (Albert Einstein)

Introduction

The decrease of the bee population in the world is an ecological disaster and a terrible threat for the earth's ecosystem. This disappearance can be explained by different factors. Indeed, the presence of "bee killers" or Asian hornets is the major problem today. However, this problem is only one of the many threats since cold and humidity are also risky situations for our bees. We can also point out the theft of hives, which has been a growing problem for beekeepers in recent years. Finally, there are artificial threats brought by human activity. First of all, the presence of chemical fertilizers and the use of insecticides have a negative impact and poison the bees. These chemical aggressions have caused their disappearance in some geographical areas

The technological product presented here is an embedded system. Its purpose is to follow in real time constants from data collected within a hive. The objective is to be able to intervene if necessary, to anticipate the problems and thus to avoid the catastrophes. With this device, we want to avoid two major concerns :

• the loss of a part of the harvest for the beekeeper and thus a decrease of his income;
• and especially the death of many individuals in a brutal way. This second danger is catastrophic from an ecological point of view.

It is at this level that we think that our system has all its utility.

Our first goal is to allow the beekeeper to control the hives remotely. For this, our system has a remote communication module. It is thus possible for the user to remotely follow the state of his different hives.

The beekeeper saves time (precious time in case of problem) and he can retrieve the data without disturbing his bees by opening the hive unexpectedly.

We would like to remind you that the main purpose of this project is, in our humble measure, to militate and work for the reduction of bee mortality in the hives.

Constraints

To allow the effective realization of the project, a certain number of constraints must be identified. The goal is to guarantee an accurate and reliable data collection for decision support.

• The taking of temperature inside the hive with a temperature accuracy of 0.1 °C
• Temperature measurement outside the hive with a temperature accuracy of 0.1 °C
• Taking the humidity inside the hive with a humidity accuracy of 2%.
• Humidity measurement outside the hive with a humidity accuracy of 2%.
• The weight of the hive with an accuracy of 100g
• The importance of the luminosity, the choice of the solar panels to conceive an autonomous system in Energy
• Designing a system that must be able to transmit data

To answer these constraints, we used :

• 3 waterproof probes for indoor temperature measurement (DS1820)
• 1 indoor humidity sensor (DHT22, whose temperature sensor is not used)
• 1 weight sensor for mass measurement placed under the hive (HX711)
• 1 single module for measuring outdoor humidity and temperature (DHT22)

Microcontroller

For the microcontroller where we put our main code, we chose the Arduino nano 33 BLE sense board for its low power consumption and its deep sleep mode included allowing a low consumption. The deep sleep mode allows to put the board in standby mode in a simple way and thus to save energy. This question of consumption is at the center of our thoughts and our choices as you can see.

Indoor Temperature and Humidity

Several sensors were considered (see above). A work of analysis and comparison was carried out. In the end, to recover data inside the hive, we chose water resistance. We selected waterproof sensors because the environment, the hive, is an environment that we describe as quite aggressive. We are in a beehive, in the middle of bees that can cover these sensors by their work, hence the need for their robustness and their ability to be waterproof.

The 3 waterproof probes for the indoor temperature sensor (DS1820) and the indoor humidity sensor (DHT22) operate in Onewire protocol on a bus. For the temperature, the three sensors are on three different buses. In fact, we use more pins on the microcontroller and therefore more connections, but we guarantee a higher reliability of data collection.

The field tests have proven a real efficiency and a high precision of the sensors we have selected.

Outdoor Temperature and Humidity

For the exterior, the sensors are subject to other constraints related to weather conditions but the problems are different than for the interior of the hive. We considered that we could go for mid-range sensors with a lower cost. In parallel, we have protected these outdoor sensors (temperature and humidity with the DHT22) with a box that we have printed specially:

Mass measurement

Weight is another essential piece of data for the beekeeper in monitoring the health of the hive. It indicates the vitality of the swarm of bees. A sudden drop in weight can indicate disappearance, death, swarming or flight of the bees. The weight sensor was imposed in this project. It is used with an Analog-Digital converter HX711.

During the various tests, the weight measurement was improved beyond the specifications. Indeed, we have succeeded in increasing the accuracy to within 16 grams. The consequence is a more precise vision of the evolution of the weight as the graph shows it:

Data exchange

The strong axis of this project is the question of remote communication. The choice to communicate with the outside world the data collected from the embedded system was to use the sigfox technology (also called LPWAN or Low Power Wide Area network). The data are sent on a minimum tempo of 10 minutes, tempo defined by the code that we have developed. However, depending on the battery and its autonomy, this frequency can change and it can go from 10 to 30 minutes. We have therefore within the code imagined this modification that puts in relation to battery autonomy and tempo of sending data.

Sigfox is a technology based on the 0G network which is a low frequency mode. This mode allows to be located in a low consumption when sending data however the weak point is located at the level of the quantity of data transmitted at each sending. This quantity is limited to 12 bytes. We have also worked on the fact of concatenating data to be sent in order to reduce their numerical weight.

Here is the data frame for sending a message to the sigfox back end:

The sigfox backend allows to collect all the data sent from the antenna according to the set up frame. On the other hand, Ubidot is the platform on which we put the data received on the backend. This platform offers an ergonomics adapted to non-specialist users. It is easy to use online. The received data have been formatted through different graphs. Each graph allows to visualize the data and their evolution in time according to the scales chosen by the user. For this question of data, their transmissions and their visualizations, our goal was to have an answer adapted to the situation and to the context of the world of bees (hives to be monitored and beekeepers who monitor the application).

So that beekeepers do not have to constantly monitor the data received, we used an option present on Ubidot which allows to send alerts by email according to the values taken by the variables associated with the measurements made by the sensors. Thus, the beekeeper will be directly alerted if the weight of his hive suddenly drops to 0kg, which means that his hive has been stolen, or if the temperature inside the hive is outside the recommended temperature range for bees.

Circuit Board (PCB)

In order to reduce the space requirement, to eliminate the use of cables and to make the system as clean as possible, a PCB was made during the project. The aim is always the same: to take up as little space as possible. Here is the PCB used:

Autonomy and power supply

The autonomy is one of the central points of the project. Our battery has a capacity of 1050 mAh. All the reflection is based on this capacity and it was necessary to work on an optimization of the consumption. It is essential that the embedded system needs to be recharged as little as possible. For this optimization, we proceeded to put the weight sensor on standby during the inactive phases of the embedded system. The DHT22 and DS1820 also automatically switch to standby.

It is thus necessary to have a consumption as low as possible and also to have a means of recharging the battery regularly. We have therefore chosen to use solar energy via a 2W solar panel for the recharging side of the system:

With this set, it is possible to ensure a stable charge and discharge cycle and prevent the system from shutting down. This can be seen on the January 2022 tests:

At the level of consumption, we are at 5 mAh. That allows to have a theoretical autonomy of a little more than 8 days without the slightest recharge (T=1050 mAh / 5 mAh in hours). All these data show the validity of the system in terms of autonomy.

To adapt the voltages of the solar panel for our battery, we use the LipoRider pro card. This card allows to deliver the right voltage to the battery. In addition to supply our PCB, we use DC / DC converter with low loss to deliver 3.3 V to the microcontroller Arduino Nano BLE sense.

System Assembly

The embedded system is placed in a high-risk environment in terms of climate. The object can be subjected to cold and humidity depending on the geographical location. We have therefore placed the whole system with the battery in a waterproof box. The connecting cables that come out of this box are fitted with special end caps adapted to this climate. The goal is to offer a technical solution that is resistant and durable over time. The beekeeper needs a simple, accessible technological answer and reliable data over a long time.