Comparative analysis of IoT capabilities, unmanned aerial vehicles and robotic technologies and identification of options for their integrated use

In modern world, digital, IoT-based and robotic systems are increasingly used in agriculture.

The scientific tasks in work package 1.4 present a comparative analysis of the capabilities of IoT, unmanned aerial vehicles and robotic technologies for the study of greenhouse gases, collection of primary information on open and closed farms and its processing in real time.

The scientific tasks in details are the following:
– Study of the amount of greenhouse gases in the cultivation of different crops by conventional method and by robotic technologies and unmanned aerial vehicles.
– Research and comparative analysis of the applicability of stationary sensor networks and mobile devices (drones, robots) for collecting primary information – open (fields) and indoor (greenhouses) agricultural areas, soil, water, air and plant biodiversity.
– Collection, merging and processing of data (Data mining, Data fusion and Data processing) from stationary and mobile sensor kits for monitoring of soils and crops in real time or in the real time dimension (Real Time or Near Real Time).

The work package provides for the determination of the total gases generated in the cultivation of certain vegetables grown by the conventional method. The total generated greenhouse gases in the cultivation of some types of field crops (wheat and corn) will be determined according to a developed methodology. It is planned to determine by how much the generated greenhouse gases will be reduced when using robotic technologies, compared to the available conventional technologies. The experience will be one-factor, as a factor will be agricultural crops (wheat, barley and corn for grain), grown by certain technologies and systems for tillage. In addition, the accompanying climatic (air temperature, relative humidity and atmospheric pressure) and soil (soil temperature and soil moisture in the layer 0-10 cm) indicators will be additionally measured by conventional methods and apparatus. These main and accompanying indicators will also be determined by synchronous observations, through developed robotic technologies and unmanned aerial vehicles.

The agrochemical indicators of the soil that will be measured are:
– dry matter and moisture;
– total organic carbon;
– ammonium and nitrate nitrogen;
– mobile phosphorus and potassium.

Humidity, lighting, heat radiation, temperature and pressure will be determined indoors (greenhouses) by means of a combined device. The average daily, maximum and minimum values of the indicators will be determined by means of hygrographs, thermographs and barographs. Automatic gas analyzers will be used in indoor air (greenhouses) to determine chemical agents and an express colorimetric method will be used to measure the concentration of chemical agents in the air.

It is planned to study the possibilities for converting data coming from stationary sensor networks in a form convenient for giving commands to a developed agricultural robot. The possibility of its use in open and indoor areas will be determined. At the stage, an opportunity for primary data collection will be provided. The possibility for data transfer to a nearby stationary storage will also be determined.

Will be analyzed:
– methods and approaches when using unmanned aerial systems for remote sensing of the earth’s surface with conventional and multispectral optical sensors for extracting primary information;
– methods and approaches for building and developing a cloud-based multilayer IoT platform, based on LoraWAN communication infrastructure, for data collection, processing, analysis and visualization;
– the reliability and possibilities for complex data extraction from air, ground robotic systems and stationary sensor stations;
– optical and conventional methods and approaches for collecting and processing informative parameters for assessment of soil, agricultural areas and plant biodiversity using a stationary sensor network based on a cloud multilayer IoT platform and LoraWAN communication infrastructure.
– the possibilities for digitalization of the process for express remote monitoring of soils in agriculture through sensor networks and mobile devices.

A stationary sensor network based on a cloud multilayer IoT platform and a LoraWAN communication infrastructure for assessing soil parameters will be built. The primary information (color – quality indicators) of the soil will be modeled through mathematical models and regression equations. An experimental study and comparative analysis of the obtained dependences for indirect evaluation of soil with reference ones will be performed.

The possibilities for applying the techniques of Artificial Intelligence and neural networks to detect correlation in independent variables will be explored in order not only to extract data, as in traditional DBMS queries, but also to extract knowledge from them, hypotheses and generate opportunities for future forecasts; as well as algorithms in the field of machine learning (observed and unobserved) – classification, clustering, regression analysis, etc., to detect correlations and generate hypotheses.