Hardware and software system for adaptive precision agriculture management within the consolidation of modern agrotechnologies in the crop production sector of the Kyrgyz Republic

Abstract

Background:

Kyrgyzstan’s economy depends on agriculture. Of strategic importance are the natural conditions that form various ecosystems in different regions. The article highlights the problems that can be addressed by developing a unified software and hardware complex that ensures the fulfillment of the tasks associated with the agricultural production cycle.

Materials and methods:

The comprehensive application of systems analysis, system engineering, and object-oriented design methods enabled the development of models for the software and hardware complex. These models were used during the development phase following the incremental life cycle model.

Results:

The article presents models of the software and hardware structure, detailing the characteristics of each component and the categories of users along with their functionalities. Examples of user tasks and corresponding scenarios for using the digital system have been created. A prototype user interface has also been developed.

Conclusion:

The use of this development fosters the creation of favorable conditions for crop cultivation in accordance with modern agricultural technology principles, allowing for ongoing and objective monitoring of plant health and farm facilities. The results can be used to create a national digital agri-food system that promotes sustainable development and food security.

1 Introduction

The current stage of societal development is characterized by the active digital transformation of processes across all areas of human activity (Smirnov and Antropova, 2022; Alexandrov et al., 2018; Babenko, 2022). Despite this, a key factor determining a country’s socio-economic condition remains its level of food security, as consistent availability of food is essential to human life (Zholbolduyeva et al., 2024). Achieving this requires an efficient agricultural production system that supplies both agricultural raw materials and healthy (environmentally clean) products.

Agriculture in the Kyrgyz Republic is characterized by an extensive mode of production, a significant share in the GDP structure, a large proportion of the population employed in the sector, a predominance of small-scale production structures, a low level of investment and technical support, and insufficient state support for the industry (Zholbolduyeva et al., 2024).

Agricultural production processes are significantly influenced by the country’s natural and climatic conditions, which feature a sharply continental climate [the southern part of the country is located in a subtropical zone, while the northern part has a continental climate (Shpedt and Aksenova, 2021)]. Another important factor shaping the development of the agricultural sector is the presence of mountainous terrain that covers nearly 75% of the national territory. This diversity of natural and climatic conditions allows for the cultivation of both heat-loving and frost-resistant crops (Parpieva et al., 2023; Shaiyldaeva et al., 2024).

Crop production in the Kyrgyz Republic includes the cultivation of grain crops (wheat, barley, oats, rye), cotton, tobacco, sugar beet, oilseeds, potatoes, vegetables, melons, forage crops, and fruit and berry crops (Karbekova and Karbekova, 2020; Agriculture, 2025). The sowing area for each of these categories is inconsistent and depends not only on product demand but also on the overall state of the industry (Koshueva and Zhumabaeva, 2023; Bektenova and Dzhumakova, 2022). Government policies have not resolved existing problems such as low labor productivity, poor technical equipment, underdeveloped logistics, and inadequate storage, processing, and delivery systems for agricultural products, as well as soil degradation and low-quality seed material (Mirlanbek et al., 2023). As a result, the least costly grain crops are cultivated most actively, including those used as livestock feed, while the more profitable technical crops, vegetables, and melons occupy only a small share of the total sown area (Agriculture, 2025; Projects, 2025). Potato production in the Kyrgyz Republic remains relatively stable, as it is a traditional crop for local farmers and has a steady domestic market (Karbekova and Karbekova, 2020; Bektenova and Dzhumakova, 2022).

All these factors negatively affect agrotechnical production parameters, leading producers to supply mainly non-standardized products that cannot meet uniform requirements for effective processing and export (Zholbolduyeva et al., 2024; Semenov et al., 2025). It should be noted that there are local examples of resource consolidation among agricultural producers, providing vertical integration with processing industries and targeting specific external markets (Zhang et al., 2023; Saparova et al., 2024). However, in most cases, such producers face difficulties competing in the domestic market with imported goods (for example, from Uzbekistan and Turkey), as the initial stages of agricultural enterprise development require substantial investment (Musarova and Adamkulova, 2023; Bakalbaeva, 2024). The cultivation of crops also requires specialized machinery and skilled personnel to implement modern agrotechnological and farm management practices (Rayhana et al., 2021).

Work in this direction is underway in the Kyrgyz Republic (Kozhomkulova et al., 2023; Semenov and Semenov, 2022). Agricultural development concepts being designed and implemented in crop production aim to reduce dependence on climatic conditions and adverse weather events by promoting greenhouse farming. Currently, there are approximately 2,000 operational greenhouses in the Kyrgyz Republic, covering more than 200 hectares; these are mostly located in the Chuy, Osh, Batken, and Jalal-Abad regions (Karbekova and Karbekova, 2020). Greenhouse environments create artificial microclimates suitable for specific crop types, enabling year-round cultivation (Volkova, 2021). Their productivity can significantly exceed that of open-field cultivation, achieved through climate control systems and protection from external diseases and pests. However, market realities and high energy costs restrict most greenhouses to small sizes. Greenhouse construction and operation widely employ technologies from Russia, Turkey, South Korea, and China (Karbekova and Karbekova, 2020).

Thus, ensuring national food security and reducing operational costs in crop production requires the integrated use of modern hardware and software tools to create a unified ecosystem for monitoring, controlling, and managing efficient plant cultivation processes. Based on this, the objective of this work is to develop a technical design for a software product that consolidates modern hardware and software solutions (including those using artificial intelligence technologies) to ensure high agricultural productivity. To achieve this objective, it is necessary to

• Determine the key characteristics defining the state of crop production in the agricultural sector of the Kyrgyz Republic;

• Establish the hardware and software structure and functional capabilities of the software product;

• Identify key user categories and define their access to relevant functions;

• Develop a prototype of a user interface that provides access to the resources and functions of the developed software product.

The study hypothesizes that the adaptability of the software–hardware complex depends on accounting for regional factors.

The scientific novelty of this work lies in developing a comprehensive model for the digital transformation of crop production processes, tailored to the specific conditions of Kyrgyzstan. The research considers the unique weather and climatic features affecting the yield of agricultural plants, as well as the cultivation practices of local farmers. This enables precise accounting of risk factors, increased efficiency in crop cultivation, and the systematization and unification of production processes in agricultural organizations through the implementation and operation of an innovative digital solution. The differentiated approach established at its core ensures flexibility and precision in applying modern technologies, depending on the specific conditions within the farming ecosystem. The scientific novelty of the work is confirmed by its theoretical and practical significance, demonstrating the uniqueness of the conducted research.

The theoretical significance of this work involves the systematization and unification of processes and characteristics that influence crop yield. The models developed in this research can serve as a foundation for further scientific studies aimed at optimizing resources for crop cultivation, forecasting changes in agroecosystem states, and integrating innovative technologies into the agricultural sector.

The practical significance of the study is the creation of an ecosystem for farmers, manufacturers of agricultural machinery and software, service providers, and government authorities. The core functions of this ecosystem include

• Real-time monitoring and control of cultivation processes;

• Consolidation of hardware, software, and human resources to efficiently organize agricultural activities;

• Facilitation of product sales.

The functional capabilities of the software product enable effective resource allocation for process and risk management through an intuitive interface. Consequently, it becomes possible to manage crop yields more precisely and mitigate negative environmental impacts. Collectively, these factors contribute positively to ecological sustainability and align with modern principles of sustainable agriculture.

The results obtained during functional and non-functional testing, as well as a case study in real agricultural conditions, proved the advantages of integrating software and hardware tools into the crop cultivation process compared to traditional methods used by Kyrgyz farmers. The author’s concept of the digital system adapts global digital technologies to the context of the Kyrgyz Republic and possesses a competitive advantage through its ability to operate under conditions of unstable connectivity and integrate with local data. Thus, the research incorporates an algorithm for the adaptive management of precision farming processes tailored to specific agroclimatic conditions.

The findings of the conducted research are systematically organized and presented across several sections:

• Introduction: The current state of the agricultural transformation process in Kyrgyzstan is outlined, forming the basis for defining the problem, objectives, and hypothesis of the study. The scientific novelty, as well as the theoretical and practical significance of the achieved results, are described.

• Literature Review: An analysis of various information sources regarding the current state of research and developments in agriculture is presented. The identified trends are discussed, considering temporal factors, the level of economic development of countries, and other factors influencing the rate of emergence and adoption of new methodologies and technologies.

• Materials and Methods: The object and subject of the study are specified, along with the methodological framework employed.

• Results: The main results derived from the application of the stated methodology are presented. These include essential characteristics of the studied object and subject necessary for developing conceptual models. A software–hardware model has been developed, detailing the minimum infrastructure requirements for ensuring the proper functioning of the digital system. Based on this, a user access model describing the available functionalities for each user category is created. Fragments of operational scenarios for the digital system and the results of its experimental validation under conditions close to real-life are also provided.

• Discussion: An analysis of the results in the context of global achievements in agricultural digitization is conducted. The limitations of the study’s findings and potential risks associated with project implementation and development, in light of Kyrgyzstan’s level of economic development, are identified.

• Conclusion: The final conclusions summarize the completed work and position the study’s contributions within the broader landscape of global advancements in digital technologies for agriculture.

2 Literature review

Agriculture is characterized by long production cycles and exposure to natural risks, along with other factors that can lead to crop losses at any stage—from cultivation and harvesting to storage and transportation. The sector also faces challenges, such as the inability to automate biological processes and the lack of productivity growth and innovation (Grankina and Vasilyev, 2024; Bondar and Logachev, 2025).

2.1 Transformation of research and development in global agriculture

Agricultural development generally follows several key directions: first, the improvement of cultivation technologies (e.g., enhancing seed quality and breeding new varieties), and second, the introduction of tools and technologies that directly impact the production processes of final goods (Da Silveira et al., 2021). The pace of change in the latter direction is closely tied to the evolution of hardware and software, a trend characteristic of global agriculture as a whole (Meshram et al., 2021).

With the advancement of information technology, software tools began to be introduced into the agricultural sector, initially to manage financial and reporting activities on farms. Early research focused on increasing agricultural productivity by creating unified information resources based on professional knowledge and by consolidating advancements from agricultural science and production (Chernysheva et al., 2022). The resulting information systems were envisioned as the final stage of integrating scientific research and technological development within a comprehensive agro-industrial cluster. Subsequently, information technology was increasingly applied to mathematical modeling of production processes and agro-monitoring in precision farming. Simultaneously, the development of GIS technologies facilitated the use of geospatial data for designing adaptive, landscape-based farming systems (Ghosh and Kumpatla, 2022; Raihan, 2024).

The subsequent evolution of technology and the growing interest of technology companies facilitated the development and implementation of hardware and software solutions that allow for the monitoring and control of crop or livestock production cycles using smart devices and unified communication channels connecting them with external stakeholders (Da Silveira et al., 2021; Lin et al., 2024). These devices recorded, transmitted, and, in some cases, processed real-time parameters of objects and their environments. As a result, a digital ecosystem emerged that enabled hardware and software tools to assist workers in performing specific agricultural tasks. Furthermore, large farms and agro-industrial complexes began to actively adopt automated tractors, combines, and other machinery equipped with autopilot systems to improve the accuracy of field operations and reduce costs for fuel, fertilizers, and other resources. Robotic systems were developed to manage entire agricultural cycles, streamlining labor-intensive processes, while unmanned aerial vehicles (drones) were used to monitor field conditions and perform remote tasks, such as targeted treatment of pest-affected crops (Volkova, 2021; Bhat and Huang, 2021; Kujawa and Niedbała, 2021).

Currently, progress in agriculture is driven by advances in artificial intelligence and machine learning (Bhat and Huang, 2021). The primary catalyst for the digital transformation of the agricultural sector was the COVID-19 pandemic, which imposed quarantines and social restrictions, thereby accelerating the development of software solutions for remote monitoring, data analytics, and process automation (Sridhar et al., 2023; Haggag, 2021). Contemporary industry research and development now integrates achievements in the Internet of Things, robotics, artificial intelligence, and geospatial analytics. Their primary objective is to support data-driven decision-making for farm management and to mitigate the negative environmental impact of agricultural activities (Bondar and Logachev, 2025; Meshram et al., 2021; Pandey and Pandey, 2023).

When assessing market trends in the agrotechnology sector, it is evident that global economic instability has led investors to favor finance projects with strong market potential and stable current positions. Consequently, many startups have declared bankruptcy or scaled back their operations (Agriculture Dive, 2025). This unstable global economic situation has also contributed to a decline in the active labor force, which, in turn, has spurred significant investment in projects involving robotic technologies and related software solutions (AgFunder, 2025). Moreover, projects addressing food security, climate change, and other global challenges are actively supported (Da Silveira et al., 2021; Agriculture Dive, 2025).

2.2 Characteristics of digital agriculture worldwide

The contemporary phase of agricultural development involves the integration of cutting-edge technologies and innovations aimed at enhancing both the efficiency and sustainability of agricultural production (Smirnov and Antropova, 2022; Da Silveira et al., 2021). Terms such as Smart Farming, Precision Agriculture, and Digital Farming have emerged to describe this new paradigm. These concepts reflect the principles of Industry 4.0, which signifies the fourth industrial revolution characterized by the digitization and automation of manufacturing processes (Shcherbakova et al., 2024).

A review of relevant literature has identified key factors affecting smart agriculture practices, which are summarized in Table 1.

Farmers typically learn about advances in Agriculture 4.0 primarily through media outlets (such as specialized journals or websites), forums, fairs, or directly from manufacturers of technical equipment and specialized software (BKT, 2025). Consequently, even within a single region, the adoption of technological advancements varies significantly across farming operations of different scales.

Overall, the level of agricultural digitization reflects a country’s economic development. Developed countries—including the United States, Germany, Canada, and Switzerland—are known for intensive production, market orientation, high labor productivity, robust governmental policies, and substantial subsidies. They benefit from integrated solutions that connect diverse sectors, from processing and logistics to machinery and inputs (Shcherbakova et al., 2024; Spielman et al., 2021; Nowak, 2021).

In contrast, transitional economies (e.g., Poland, the Czech Republic, Russia, China, and Uzbekistan) exhibit varying stages of agricultural transformation. For instance, while countries like China and Russia have made significant progress toward self-sufficiency and food security, others face challenges due to insufficient investment, structural reforms, privatization, and intense competition from more established foreign competitors (Zholbolduyeva et al., 2024; Shpedt and Aksenova, 2021; Berg et al., 2024).

Developing nations, particularly in Eastern Europe, South Asia, the Middle East, North Africa, Latin America, and the Caribbean, rely heavily on extensive forms of agriculture geared toward local consumption rather than export (Da Silveira et al., 2021; Lin et al., 2024). Their challenges include inadequate industrial development, weak infrastructure, and outdated techniques. Many farmers in these regions depend largely on traditional knowledge and are often hesitant to adopt modern research findings or invest in improved seeds (Pryazhnikova, 2023). However, given rising domestic demands driven by population growth, they must increase output by incorporating new technologies (Lin et al., 2024; Pandey et al., 2022).

Least-developed countries face severe limitations in accessing advanced technologies, adequate infrastructure, and basic education, making widespread adoption of digital agriculture nearly impossible (Norton et al., 2021). Nonetheless, some simpler software applications, such as weather updates or peer-to-peer networking, remain accessible options for marginalized communities (Pandey et al., 2022).

As an example, an excerpt from the list of analyzed software and hardware solutions in the field of global digital agriculture is presented in Table 2.

The data indicate a wide range of software and hardware solutions currently employed in agriculture. Each developer offers a unique set of functional capabilities alongside the necessary equipment for full functionality. The effectiveness of these digital systems varies depending on the region’s or country’s level of agricultural development, as well as the prevalence of specific crops.

Regardless of their economic circumstances or stage of digital evolution, farmers worldwide cite interoperability issues between disparate digital systems and challenges in synchronizing products from multiple vendors as critical obstacles to further advancement (BKT, 2025).

Research confirms that these factors significantly influence farmers’ interest in adopting such technologies. The insights gained from analyzing these software advancements form the basis for developing a concept for a software–hardware complex tailored to the conditions of Kyrgyzstan.

Amid these global trends, the agricultural sector of the Kyrgyz Republic is focused on ensuring the population’s access to high-quality food by transforming the industry into a supplier of premium, environmentally clean, and organic products (Zholbolduyeva et al., 2024; Abdiev et al., 2024). The National Development Strategy, adopted through 2040, mandates state support for large agricultural enterprises, farms, and cooperatives involved in the production and processing of high-value agricultural products. Achieving these objectives requires the comprehensive implementation of modern hardware and software tools to ensure efficiency across the entire cycle of crop cultivation, storage, transportation, and distribution. To this end, the state enterprise “Agro-Smart” has been established to enable the digitalization of the country’s agro-industrial complex by centralizing information flows and providing farmers with access to a unified data resource base (Projects, 2025).

3 Materials and methods

The object of this study is the crop production sector of the Kyrgyz Republic. The study focuses on the management of crop cultivation.

The creation of a software product is a complex process comprising multiple stages, defined by the specific tasks and the nature of the work involved. In implementing this project, an incremental software life cycle model was adopted. This model involves implementing functions in steps, producing a functional version of the software product at the end of each phase. This methodology is widely applied in the development of software systems where requirements are clearly defined, the system decomposition is well understood, and there are significant risks in the development process, such as changes in problem domain components, technologies, or tools (Kyeremeh, 2021). This approach enabled a flexible software development process, allowing for modifications in each iteration and facilitating rapid testing and debugging.

The rigor of the chosen model necessitates a meticulously elaborated design phase, as software architecture requirements must be formulated and all system components identified and structured. To achieve this, the following general scientific methods were applied sequentially:

• The systems analysis method was employed to study the problem domain, which encompasses the set of objects, processes, and external factors influencing the condition of plants cultivated in greenhouse environments. As noted in previous research, applying this method enables the identification of key characteristics that subsequently determine the features of digital transformation (Projects, 2025; Semenov et al., 2025). To implement this approach, a comprehensive set of techniques was applied, including decomposition, stepwise refinement, abstraction, systematization, statistical analysis, structuring, and synthesis.

• The graphical method was used to visually represent quantitative and qualitative data obtained through systems analysis. This method is widely applied for visualizing large volumes of data, identifying patterns and trends, and comparing different groups of objects (Grankina and Vasilyev, 2024; Logachev, 2024). In this study, the method was used to create bar charts (to compare quantitative indicators of different object characteristics), pie charts (to determine the share of each component relative to the total volume), as well as structural and information technology diagrams (to display data flows, information interactions, and communications across transmission channels, network nodes, servers, and other components of the system’s hierarchical structure).

• Object-oriented design was utilized for the formalization of user functionality. This method is based on the concept of integrating data and the methods that operate on them within a single object, which can interact with other objects through defined interfaces (Bondar and Logachev, 2025; Orlik and Krasnikova, 2024). The method is widely used in software development and is based on the principles of abstraction, encapsulation, inheritance, and polymorphism.

• The systems engineering method represents an interdisciplinary approach to the design, development, and management of complex systems throughout all stages of the software product life cycle (Buede and Miller, 2024). In this study, an integrated approach encompassing software, hardware, processes, and human resources is applied at all stages of project implementation, ensuring alignment with the project’s objectives, plans, and methodologies.

• Methods for processing data from Internet of Things devices. The developed software–hardware architecture involves the integration of various devices, which necessitates a comprehensive approach to managing large volumes of heterogeneous data. The work employed the most common methods, which can be grouped into categories: collection of raw data, preprocessing, analytics and modeling, data visualization and presentation, and optimization and quality improvement (Guerrero-Ulloa et al., 2023; Fortino et al., 2020; Kour and Arora, 2020; Zhang et al., 2025). Data obtained directly from sensors or devices required normalization of values (including the removal of noise and anomalies) and the creation of unified aggregated metrics before further use in the digital system.

• Techniques for Building Neural Models. A set of methods enables the acquisition of an architecture (network type, layers), data preparation, and model training. The key stages of design included determining the network type, selecting the sequence of layers, regularization, and weight initialization. According to research, convolutional networks were selected for neural networks using images, and recurrent networks for sequences (Taye, 2023; Yang et al., 2023; Qamar and Zardari, 2023). To increase the speed and quality of training, weight coefficients and regularization techniques were used to prevent overfitting.

4 Results

4.1 Key characteristics of the research object

Agriculture is a vital sector in the Kyrgyz Republic, employing a significant portion of the working-age population. While livestock and crop production primarily meet domestic needs, surplus products are exported. The main agricultural producers include state-owned, collective, and private farms, as well as individual entrepreneurs. Figure 1 shows their total number, distributed across the country’s regions.

Figure 1

An analysis of the literature reveals a clear trend in the Kyrgyz Republic: a decrease in the number of state-owned and collective farms and an increase in individual entrepreneurs (Zholbolduyeva et al., 2024; Agriculture, 2025; Projects, 2025). Figure 2 shows the percentage distribution of farms by type. The values in parentheses indicate the growth rate compared to the previous year.

Figure 2

The physical volume index of agricultural production for the crop sector was 98.4% of the previous year’s figure (Projects, 2025). Agricultural producers cultivate a variety of crops, including cereals, legumes, potatoes, vegetables, cotton, tobacco, melons, oilseeds, sugar beets, fruits, berries, grapes, and forage crops. The distribution of these cultivated plants by region is shown in Figure 3.

Figure 3

Developing a digital system to monitor and control plant cultivation requires a detailed characterization of each entity involved in the process. Each crop type has specific requirements for its ecosystem, cultivation practices, and life cycle duration. The collection, systematization, and formalization of this data are essential for generating accurate recommendations and forecasts for high-yield cultivation. This information is fundamental for creating the models that enable the artificial intelligence within the digital system to function.

To forecast trends for the crop production sector as a whole, data on each crop type were required. Figures 4–7 present the gross output of each agricultural enterprise in the Kyrgyz Republic for the most recent year. The labels for each crop category indicate total output and growth rates compared to the previous year (in percent).

Figure 4

Figure 5

Figure 6

Figure 7

The hardware and software structure of the digital system was designed with the current technical capacity of agricultural enterprises in mind. Uninterrupted access to the system’s functions and resources requires both a local network and Internet connectivity. Figure 8 presents official statistics on the distribution of Internet access points across the regions of the Kyrgyz Republic. Figure 9 shows the number of local computer networks in use. Figure 10 illustrates the number of enterprises utilizing information and communication technologies (ICT) for business operations.

Figure 8

Figure 9

Figure 10

The analysis reveals that the adoption of hardware and software in agricultural operations remains limited. However, a positive trend indicates a gradual improvement in these metrics. This suggests rising digital literacy, creating an opportunity to effectively implement specialized farm management tools, provided the necessary technical upgrades are made.

Individual entrepreneurs, who often operate small-scale farms, require special consideration. Such farms need a tailored approach to digital transformation due to their limited resources. Consequently, this study accounted for the possibility of farm management via pre-installed software capable of autonomous operation. This approach requires access to a personal computer, which, according to our analysis, is available to a significant portion of the population (Figure 11).

Figure 11

The analysis of agricultural technical resources shows that the primary machinery for crop cultivation includes tractors, combines, seeders, soil cultivators, and irrigation systems (Shpedt and Aksenova, 2021; Parpieva et al., 2023). A critical factor for agriculture in the Kyrgyz Republic is the limited availability of water resources for irrigation. Therefore, when defining the digital system’s functional capabilities, particular emphasis was placed on integrating efficient irrigation systems to optimize water consumption and ensure the rational use of moisture.

4.2 Architecture and technical specifications of the digital system

The analysis of the problem domain showed that the software product should be based on a client–server architecture. This architecture enables interaction between users and the digital system’s services, which perform specific computational tasks. Furthermore, it allows for the integration of various technical devices operating on the farm, either under human control or through artificial intelligence (e.g., robotic systems and unmanned aerial vehicles).

The system’s hardware configuration may vary depending on the type of farming activity and the range of equipment used in agricultural operations, while the list of software modules remains consistent. Figure 12 presents the hardware and software architecture of the digital system for a crop production farm.

Figure 12

A typical crop production farm occupies a large area, often located far from major population centers. Consequently, maintaining a stable internet connection to ensure continuous system operation can be challenging. To mitigate this issue, the system architecture supports autonomous operation. For this purpose, the farm requires a facility equipped with a personal computer that aggregates all data concerning the current state of the monitored objects.

Table 3 describes each component of the developed hardware and software architecture.

The hardware components may vary depending on the type of crop farm. Plants may be cultivated either in greenhouses or in open fields, necessitating the use of different sensor types for effective monitoring. These can include wired sensors, Wi-Fi-equipped sensors, or sensors mounted on mobile devices.

Sensors installed in greenhouses must periodically record measurements of the microclimate, such as air humidity, illumination, and temperature, as well as soil conditions, including moisture, acidity, and density. The integration of video and audio data from corresponding devices enables further analysis to identify various issues. For instance, video or photo recordings can help detect poor-quality work by employees, unauthorized access, or allow for the assessment of plant health.

Sound analysis provides additional critical data. It can detect noise pollution that may negatively affect plant growth, identify abnormal situations caused by malfunctioning equipment, or even recognize structural damage to facilities. Furthermore, in greenhouse farms that utilize bees for pollination, sound analysis can be applied to monitor the health and activity of bee colonies.

These principles of monitoring and control can also be adapted for open-field environments. This requires the digital system to be properly configured and synchronized with the installed sensors, detectors, and other technical devices. The primary challenge in this context is ensuring the reliable and timely transmission of data from the distributed measurement points back to the central personal computer.

The software–hardware architecture involves the use of microservices, which can be based on neural networks or artificial intelligence technologies. Table 4 presents fragments of a neural network implemented in the Python programming language (version 3.13), with comments for implementing the function of determining the most favorable neighborhood for agricultural crops.

The influencing factors include soil characteristics, regional climatic conditions, and the specific requirements of each crop (e.g., disease resistance and nutrient needs). The data encompass soil types, average precipitation, temperature regimes, daylight duration, crop fertilizer requirements, and pest presence. The objective is to determine the optimal spatial arrangement of plants to maximize yield metrics and minimize the risk of disease and crop damage due to unfavorable neighboring species.

Data collection involves creating a dataset containing

• Geographical location of plots,

• Soil physical properties (pH, moisture),

• Climatic condition indicators (temperature, precipitation),

• Crop composition and yield dependency on adjacent plant species.

For processing, an adjacency matrix is used to reflect the relationships between different plant species. For instance, crop A has a positive influence on crop B if its presence enhances the latter’s growth, or a negative influence if it reduces productivity. The adjacency matrix is constructed based on an analysis of results from peer-reviewed educational, methodological, and scientific–practical publications (Bogatyreva and Yarkova, 2024; Kiryushin and Kiryushin, 2023; Matyuk et al., 2021; Kotlyarova et al., 2022).

Standard normalization methods are applied for data transformation. To improve prediction quality, one-hot encoding is introduced for categorical features (soil types, cultivation regions).

A brief description of the code fragments presented in Table 4 is as follows:

• Dense layers are used to compute non-linear transformations of features.

• Batch normalization normalizes activation values, accelerating training and improving model stability.

• Dropout prevents overfitting by randomly deactivating a portion of neurons.

• The final layer returns the probability of successful interaction between two specific crops (binary classification).

• The model is trained using the Adam optimizer and binary cross-entropy as the loss function, enabling efficient identification of the optimal neural network weights.

• Classification accuracy and the F1 score are evaluated to demonstrate the effectiveness of selecting the optimal combination of neighboring crops.

Here are the minimum hardware requirements for deploying client-side components in a small-scale farm setting.

Personal computer requirements:

• Processor: Quad-core Intel Core i3/i5 or AMD Ryzen 3/5 series. Multithreading support ensures sufficient performance for running multiple applications simultaneously, managing network services, and autonomously handling agricultural tasks.

• RAM: Minimum 8 GB DDR4/DDR5 RAM, expandable to 16 GB for larger farm operations. This capacity guarantees stable OS operation, smooth performance of client applications, and the virtualization of certain digital system components.

• Storage: An SSD with a minimum capacity of 256 GB, adequate for installing the operating system, necessary software, storing user settings and backups, and holding temporary data before it is transferred for intelligent processing and archival.

• Graphics card: A standard integrated GPU. An optional upgrade to a mid-range graphics card (e.g., NVIDIA GeForce GT/GTX series or AMD RX series) is available for enhanced visualization of field maps and predictive models.

• Motherboard: Must feature ports such as USB Type-C, HDMI, VGA, DVI, and Gigabit Ethernet LAN connectors. Additional PCI-E x16 slots may be useful for scaling the digital ecosystem.

• Network interface: A connection speed of at least 10 Mb/s.

• Power supply unit: An uninterruptible power supply (UPS) to ensure reliable autonomous operation and protection against voltage fluctuations typical in rural areas.

• Operating system: Windows 10/11 or Ubuntu Server 22.04 LTS.

Mobile device requirements:

• Operating system: Android 7.0 (Nougat) or later / iOS 11 or higher.

• Screen resolution: Minimum of 1,280 × 720 pixels.

• Camera: A front camera capable of capturing photos and videos for documenting object states, verifying completed work, and generating reports.

• Internal storage: Minimum of 16 GB of free internal memory.

• RAM: At least 2 GB.

• Wi-Fi connectivity: Compatible with IEEE 802.11b/g/n/ac standards.

• Bluetooth: Version 4.0 or above.

• GPS module: Integrated for precise location tracking.

• Audio output: Built-in headset jack.

• Battery capacity: At least 1800 mAh, providing sufficient battery life for daily app usage throughout working hours.

The conducted statistical data analysis demonstrates that the formulated requirements for software–hardware equipment in a small-scale farm are achievable.

It should be noted that the developed software architecture is scalable, allowing for several technical approaches to be implemented.

If necessary, vertical scaling can be employed, which involves increasing the resources of the servers in use (e.g., adding RAM or storage, and upgrading processors). This approach is suitable during the early stages of operation under conditions of moderate growth in server load.

When expanding functional capabilities, horizontal scaling should be implemented. This type of scaling involves adding new, independent nodes to the architecture of the existing digital system. The developed digital system architecture allows for virtually unlimited growth in the number of services used. This is achieved through the interaction of small, loosely coupled, and easily modifiable modules, each performing a specific function within a single process.

The separation of databases into distinct components of the software–hardware architecture enables data management through the creation of replicas for load distribution (e.g., for read operations) and clustering multiple servers to function as a single system. Furthermore, a technical approach to data management has been applied that allows for

• Centralized storage of heterogeneous data formats (utilizing databases and data warehouses);

• Data integration into a unified system through transformation after extraction from various sources;

• Ensuring quality: eliminating duplicates and continuously updating data to maintain reliability;

• Enabling continuous processing and analytics through persistent access by artificial intelligence tools or neural networks.

The developed software–hardware architecture provides the capability to implement various methods to meet information security requirements. These include

• Technical methods: performing data backups for recovery, access control (privilege separation, user identification, and authentication), antivirus protection for detecting and removing malware, cryptographic protection for data at rest and in transit, and protection against unauthorized access.

• Organizational methods: different aspects of the developed concept can be used to create instructions and regulations for working with data and the functional capabilities of the software–hardware complex, train personnel, and monitor security based on incidents.

• Physical methods: protecting the premises housing the key infrastructure elements of the software–hardware complex.

• Legal methods: compliance with the current legislation of the country where the servers are located and where the software–hardware complex is used.

4.3 Characteristics of digital system users

The digital system provides two modes of access: a mobile application and a desktop application.

The mobile application is designed for workers performing tasks directly at farm sites, such as greenhouses, fields, and storage facilities. It provides access to a notification system detailing the nature and schedule of tasks. Users can also record task completion results and document the current condition of farm assets.

Furthermore, the application enables users to report issues concerning greenhouses, fields, equipment, and cultivated plants. These reports can then be analyzed automatically or by specialized experts. This functionality digitizes the problem-reporting process and ensures the timely organization of corrective actions.

Figure 13 presents fragments of the mobile application’s user interface.

Figure 13

Task completion confirmation can be executed in two modes: manual or automatic. Automatic mode is applicable only when results can be recorded using installed sensors (e.g., for activating additional lighting, irrigation, or fertilizer application) or cameras (e.g., for mechanical soil treatment, sanitary pruning, or pest control).

The mobile application is available to users with the following roles: agronomist, horticulturist, operator, and chief engineer. Table 5 provides a brief description of their functional capabilities.

It should be noted that labor resources on a greenhouse farm may be limited. Therefore, functional capabilities are grouped to remain accessible even with minimal staff. Table 5 shows the correspondence between user roles and positions that may exist in the staffing structure of a farm, regardless of the type and quantity of cultivated products. It is important to emphasize that the management of role-based functionalities and resource accessibility is handled through standard system administrator mechanisms and scenarios.

4.4 Case study in agricultural conditions

User interaction scenarios within the digital system are defined by a set of tasks aimed at achieving specific goals for each plant type. Depending on available resources (such as consumables, workforce, and time), the following objectives may be defined:

• Achieving high yields within a specific period;

• Obtaining a sufficient yield with available resources;

• Achieving adequate productivity in a shorter period by using increased resources;

• Obtaining a feasible yield under force majeure conditions (for example, severe weather deterioration, or technical or other emergencies within the farm or surrounding area).

Based on these objectives, the system can generate the following types of tasks either manually or automatically:

• A task for the operator to achieve specified microclimate parameters in the greenhouse or to perform technical maintenance of the building and greenhouse systems.

• A task for the chief engineer to approve the purchase of fertilizers (upon request from the agronomist).

• A task for a worker to remove plants, harvest crops, or apply fertilizers.

• A task for the chief engineer to approve the schedule for launching unmanned aerial vehicles for field treatment.

Each task includes the following attributes: assignment date, completion deadline, responsible executor, and required resources.

Users in the roles of chief engineer, agronomist, and operator have access to the desktop application, which can be installed in the farm’s administrative building, as well as to the mobile application. Figure 14 presents fragments of the prototype user interface for the desktop greenhouse management application.

Figure 14

The desktop application allows users to define the system’s initial configuration parameters, including soil composition, cultivated crops, cultivation modes, types of equipment used, and personnel involved. It also enables the reception of automatically analyzed greenhouse parameters for real-time management and predictive decision-making.

To reduce the load on the farm’s infrastructure, all data and analytical results are stored in the cloud. This approach allows calculations to be performed remotely, enables configuration of the digital system, and supports the integration of new solutions without requiring upgrades to the greenhouse’s hardware infrastructure for computational processes.

The business logic of the digital system is organized so that each element of the client–server architecture has its own set of functional capabilities, implemented as separate software modules. Each module executes its own algorithms and technologies, allowing for the distribution of user load during request processing. The list of software modules includes a module for primary processing of greenhouse condition data, a module for distributing user tasks, a module for monitoring and controlling task execution, a module for monitoring and controlling greenhouse conditions, a forecasting and analytics module, and a data archiving module. Processed data are stored in a database server, while archived data are stored in a cloud repository.

An example user interaction scenario with the system’s resources and functions is as follows. A sensor records humidity data in the greenhouse and sends it to the primary data processing module, where it is converted into a format suitable for use within the digital system. The processed data are saved in the database and become available to the module responsible for monitoring and controlling greenhouse conditions. If deviations from the normative values are detected, a notification is generated, stored in the database, and made available to the task distribution module (when no automatic climate control system is installed in the greenhouse).

Based on user data, a task is created that includes the issue’s priority, deadlines, responsible personnel, and a task list. The task is then stored in the database and becomes accessible to the designated users. Notifications are sent as text and sound alerts to their devices. Depending on the importance of the issue, the greenhouse monitoring module can automatically initiate unscheduled measurements using the relevant sensors to verify whether the issue has been resolved (in cases where manual confirmation has not been received).

The user receives a notification about the new task and can view its details in the general task list. The task list is dynamic: the order of tasks is determined by their urgency and deadlines. Thus, the tasks requiring the most immediate attention appear at the top of the list. When a user manually confirms task completion, the task monitoring module processes the signal and updates the database with the corresponding task completion status.

4.5 Description of functional capabilities in the context of the digital system’s commercial implementation

The developed software product is intended for commercial use. The availability of its functions for the end user depends on the state of the material and technical resources, as well as the farmer’s willingness to implement a unified farm management system.

The architecture of the digital system allows for the integration of various devices (sensors, cameras) and equipment (drones, irrigation systems, etc.). Some of these may already be in use by the farmer. Therefore, during the implementation stage, achieving a high level of integration for such equipment is necessary to create a unified ecosystem. If specific devices or equipment are not present in the farming operation, the digital system can be provided on a “turnkey” basis, including the installation and configuration of all necessary components.

The digital system operates on a tiered subscription model. This is essential for maintaining common servers and services, as well as for compensating specialists responsible for system management and user support. The following subscription types are proposed:

• Demo: Allows viewing agricultural information without the ability to use control and monitoring functions. Available to any unauthenticated user. Free access.

• Basic: An extended version of the demo subscription. Accessible to users with a personal account, it offers limited functionality that operates only in manual mode. Free access.

• Standard: Provides users with resources and features for both manual and intelligent control of farm conditions. Available to verified farmers after payment is received.

• Professional: An extension of the Standard subscription, this includes the integration of devices and machinery into the digital system.

Table 6 presents an overview of the available functionalities based on the subscription type.

The cost of paid subscription plans is determined by the number of workers and the amount of equipment at a given agricultural enterprise. This pricing structure is justified by the need for additional computational resources to perform functions involving a large number of operational elements.

To promote the adoption of digital solutions and improve digital literacy in the agricultural sector, educational seminars and workshops may be organized for farmers. Software users will also have access to round-the-clock customer support.

The price of each subscription can vary based on several factors. Costs may be reduced through government co-financing or private investment in the project.

4.6 Performance characteristics of the digital system

At this stage of software development, the functionality of the digital system module responsible for monitoring and controlling the cultivation of vegetable seedlings in greenhouse conditions has been implemented and tested. The facility where the microclimate was created and maintained, along with its technical equipment, corresponds to the conditions of a small farm or an enterprise organized by an individual entrepreneur on a private plot of land.

During testing and pilot operation, sensors were installed and synchronized to measure and periodically transmit data on soil moisture, light intensity near cultivated plants, soil acidity, and other microclimatic indicators. The collected data were automatically processed by the digital system. Based on the results, tasks were generated to monitor and, when necessary, adjust the specified parameters.

The list of tasks and characteristics defining the scope and frequency of operations was developed based on digitized reference information on effective agricultural crop cultivation. Tasks were distributed among system users according to their assigned roles. In addition, notifications were configured to alert users about new task assignments, confirmations of receipt, and task completion.

Table 7 presents a list of performance metrics for the digital system, aligned with its validated functional capabilities.

It should be noted that the performance characteristics shown in Table 7 are relative and dependent on the type of tasks performed over a specific time interval. These values represent a snapshot at a specific moment or averaged values. Given the conceptual nature of the work performed, such values are necessary for developing the project’s strategy and creating high-load testing scenarios for the digital system as its functional capabilities expand and the volume of processed resources increases.

The user interface of the digital system was evaluated by users without prior professional experience in digital transformation processes. It was found to be user-friendly and intuitive, ensuring comfortable interaction with the system’s functionality and allowing prompt, adequate responses to notifications about deviations from standard parameters.

The use of the incremental software life cycle model enabled the development of an operational version of a system component at an early stage of the project. This working version was capable of performing the full range of functions required to collect user feedback for further development and refinement.

5 Discussion

The primary goal of research and development in agricultural technologies is to enhance the efficiency and sustainability of agriculture (Kumar et al., 2024). This is achieved by focusing on improving crop yields, enhancing product quality, and optimizing production costs. At the current stage of technological development, these objectives are addressed through solutions based on automation, digitalization, precision agriculture, and biotechnology (Meshram et al., 2021; Kujawa and Niedbała, 2021; Abdiev et al., 2024; Manida and Ganeshan, 2021). Modern concepts for agricultural modernization—encompassing crop production, animal husbandry, aquaculture, and other areas—are built on innovations that integrate artificial intelligence, the Internet of Things, unmanned aerial vehicles, and satellite monitoring with traditional agronomic methods, such as comprehensive crop care, soil treatment, and fertilizer application (Chernysheva et al., 2022; Kujawa and Niedbała, 2021; Sridhar et al., 2023). The methodology for applying agricultural technologies involves systematically addressing factors that negatively affect crop yield and, consequently, the quality of agricultural products (Pandey and Pandey, 2023; Bunkin et al., 2024). The more complex the environmental conditions and the higher the expected yield, the more diverse and specialized the required agrotechnological measures become. Therefore, each farm selects its own specific set of agricultural technologies.

5.1 Analysis of results in the context of the current state of digital transformation processes

The software product developed as part of this study is designed to create a digital ecosystem. This ecosystem includes monitoring the condition of cultivated plants using IoT sensors, applying artificial intelligence technologies to analyze microclimate parameters, generating crop care recommendations, monitoring farm operations, and managing automated systems for irrigation, watering, lighting, and temperature control. It also integrates unmanned aerial and ground vehicles for monitoring plant growth, ensuring security, and preserving farm infrastructure. All these components align with the methods and tools used in modern agrotechnology.

Researchers such as Bondar and Logachev (2025), Rayhana et al. (2021), Zhang et al. (2023), Logachev and Simonov (2024), and Grankina and Vasilyev (2024) emphasize that the use of modern machinery and digital technologies allows not only continuous monitoring of most elements involved in crop cultivation but also the generation of recommendations for operational management to achieve higher yields.

Management, including operational management, is a complex process in which a subject organizes an object’s activities to achieve specific goals (Czvetkó et al., 2022). Studies on this topic indicate that regulating such processes involves maintaining certain system parameters within predefined limits (Karbekova and Karbekova, 2020; Semenov et al., 2025; Volkova, 2021; Logachev, 2024). A variety of methods and tools can integrate hardware and software components to enable the collection of large volumes of heterogeneous data, objective analysis, controlled access, and real-time adjustment of technological operations. The developed hardware and software architecture for the digital system is designed to integrate such tools to perform these functions.

The developed concept of the digital system is based on data obtained from various devices. As noted in research related to the Internet of Things, there is an increase in resource manageability, process optimization, and improvement in the accuracy of forecasts or recommendations (AlZubi and Galyna, 2023; Mowla et al., 2023; Sharma and Shivandu, 2024). Analyzing the methodologies applied by modern researchers and developers of similar software, it can be concluded that the obtained results are comparable. Technologies for data acquisition and processing are actively used to create neural network models for software with artificial intelligence functions (Fuentes-Peñailillo et al., 2024; Gebresenbet et al., 2023; Rabhi et al., 2025; Romero-Gainza and Stewart, 2023).

To ensure efficient use of the digital system’s resources, the developers have implemented an access control policy (Kendyala et al., 2023; Pasechnikov and Avdeev, 2020). This policy governs the rights and privileges granted to system users regarding data objects and functional capabilities. A key approach to implementing access control is role-based management. This method involves designers defining user categories based on common characteristics and assigning corresponding sets of permissions and privileges. Typically, this differentiation is based on an employee’s position or job responsibilities within the organization (Bondar and Logachev, 2025; Kabakov, 2022). In the developed digital system, user roles are defined according to the functions performed by farm personnel. This principle ensures that only authorized staff can access and use system resources relevant to their professional duties, thereby preventing conflicts of interest in the allocation of materials and responsibilities.

Researchers have also noted that, depending on the form of organization, available resources, land area, and other factors, a farm may employ a limited number of workers (Climate FieldView, 2025; Zhuravleva et al., 2022; Maatova et al., 2025). Consequently, a single employee may perform multiple roles simultaneously (for example, the farm owner might also serve as the chief engineer and technologist). Therefore, the developed access control model allows for assigning multiple roles to a single individual without restricting functionality within each role.

Given the general level of digital literacy and the specific nature of farm operations, developing an intuitive and user-friendly graphical interface is essential to ensure seamless access to the system’s functions and resources. An analysis of software development practices in agriculture and other sectors shows that object-oriented design methods are widely used for creating user interfaces (Grankina and Vasilyev, 2024; Bondar and Logachev, 2025; Bodker, 2021). These methods allow developers to adhere to the core principles of object-oriented programming: abstraction (defining essential object characteristics), encapsulation (managing access to object properties and methods), inheritance (reusing properties and methods across objects), and polymorphism (using the same method to perform different actions depending on the object type) (Logachev and Simonov, 2024; Dathan and Ramnath, 2025).

As part of this project, a graphical user interface has been implemented, with control elements adapted for different device types, including smartphones, tablets, personal computers, and laptops. The resulting program code is well-structured, clear, and easily modifiable, allowing the software product’s functionality to evolve alongside project objectives.

5.2 Study limitations and regional risks associated with the operation of the digital system

Based on the factors of digital agriculture (Table 1), commercially available software–hardware solutions (Table 2), results of statistical data analysis, and the developed concept of a software–hardware complex for Kyrgyzstan, the following conclusions can be drawn:

• Minimal software and technical prerequisites exist in the country for implementing a comprehensive digital system suitable for farms of various sizes. This is supported by national development programs aimed at achieving food security in Kyrgyzstan.

• A farm’s digital ecosystem should correspond to its productive capacity. The designed system architecture allows for scalability and customizable settings. Flexible subscription models enable the tailoring of service fees and depreciation expenses.

• Existing proprietary software solutions often limit opportunities for integrating third-party developments. While our design aligns with this characteristic in some respects, the architecture still permits the integration of external modules, including those from other companies.

• Current solutions exhibit clear regional specificity: Their functional capabilities are influenced by climate, geographical position, and the unique characteristics of local crop cultivation. Our proposed digital system takes these factors into account, offering functionalities and resources tailored to the specific needs of Kyrgyzstan’s agricultural industry.

• Functional capabilities must be grounded in artificial intelligence, big data, IoT, and data analytics technologies. This requirement has been met in the presented study.

The findings align with the study’s objectives and subjects. Their validity and reliability are confirmed by the consistent application of established scientific methods and their correspondence with findings from related research.

However, the study has several limitations:

• Methodological limitations: The research focuses on crop production in the Kyrgyz Republic and accounts for regional specificities. Consequently, the results cannot be directly extrapolated to other countries without further studies that consider local agricultural conditions. The testing phase was conducted on a limited number of crops. For other crop types, additional data collection and systematization of relevant agricultural practices are required. The research was also constrained by time; in agriculture, accurate results often require multi-seasonal analysis to assess the long-term impact of applied measures on soil fertility, crop yield, adaptation to climate cycles, economic efficiency, and other time-dependent factors.

• Technical limitations: The precision of the system’s recommendations may be affected by outdated information, sensor accuracy (considering their service life), and the quality of communication channels. Furthermore, local terrain and climatic conditions can impact the performance of the connectivity infrastructure.

6 Conclusion

Digital transformation in agriculture is a global trend aimed at increasing the sector’s productivity by automating as many stages of the production cycle as possible, reducing losses, and optimizing resource management. During a single season, a farmer must make numerous decisions concerning not only the condition of cultivated plants but also threats to agricultural land and the facilities located on it. The use of software tools enables production optimization and efficient resource allocation based on data from automated microclimate control systems (for greenhouse farming), weather monitoring systems (for open-field cultivation), irrigation and lighting systems, as well as for the organization of machinery operations and spatial planning. Creating favorable conditions for plant cultivation not only ensures high crop yields but also supports environmental sustainability.

The development of integrated hardware and software systems that incorporate diverse agrotechnologies, including high-tech solutions, facilitates not only efficient farm management but also the promotion of environmentally responsible practices. This establishes a paradigm of an integrated approach that combines social responsibility, economic efficiency, and environmental safety. By comprehensively applying such technologies in crop production, all participants in the agricultural sector can ensure a sufficient food supply for the population while securing future food availability through efficient resource use. Rejecting this approach leads to intensive crop production models, which are inherently unsustainable. Such models rely on the excessive exploitation of available resources and, in the long term, increase the risk of yield losses.

Hardware and software systems that incorporate artificial intelligence and Internet of Things technologies contribute to the rational use of agricultural technologies, regardless of farm size, by relying on unified and scientifically grounded crop cultivation methods. These systems enable continuous and objective monitoring of various agrotechnical processes, systematically addressing factors that limit crop yields and product quality while maintaining soil fertility and ecosystem biodiversity.

The use of digital technologies also expands small farmers’ access to shared information, resources, financial instruments, markets, and the latest scientific and technological advancements. This enhances overall literacy in sustainable agriculture. Furthermore, it lays the foundation for a digital agri-food ecosystem that consolidates small farming enterprises, enables state control and monitoring of their activities, and ensures consumer access to locally produced agricultural goods.

References

• 1

  AbdievM. J.MatalievaN. J.ArstanbekovA. K. (2024). Modern trends in agricultural development in the Kyrgyz Republic. Econ. Bus. Theory Pract.11, 6–9. doi: 10.24412/2411-0450-2024-11-1-6-9

  • CrossRef
  • Google Scholar

• 2

  Aerobotics. (2025). Leading yield forecast technology for fruit companies. Available online at: https://aerobotics.com (Accessed September 30, 2025)

  • Google Scholar

• 3

  AgFunder (2025). AgFunder global AgriFoodTech investment report 2024. Available online at: https://agfunder.com/research/agfunder-global-agrifoodtech-investment-report-2024 (Accessed March 12, 2025)

  • Google Scholar

• 4

  Agriculture. (2025) National Statistical Committee of the Kyrgyz Republic. Available online at: https://stat.gov.kg/ru/statistics/selskoe-hozyajstvo (Accessed March 1, 2025)

  • Google Scholar

• 5

  Agriculture Dive. (2025) Investment in agtech startups plummeted in 2024. Available online at: https://www.agriculturedive.com/news/agtech-vc-deals-plummet-startups-2024/738361 (Accessed March 12, 2025)

  • Google Scholar

• 6

  AlexandrovD. A.BoukhanovskyA. V.ChugunovA. V.KabanovY.KoltsovaO. (2018). Digital transformation and global society. Cham, Switzerland: Springer Cham.

  • Google Scholar

• 7

  AlZubiA. A.GalynaK. (2023). Artificial intelligence and internet of things for sustainable farming and smart agriculture. IEEE Access11, 78686–78692. doi: 10.1109/ACCESS.2023.3298215

  • CrossRef
  • Google Scholar

• 8

  BabenkoA. G. (2022). Scientific, technological and other bases of production processes digitalization. Miner. Min. Eng.4, 90–100. doi: 10.21440/0536-1028-2022-4-90-100

  • CrossRef
  • Google Scholar

• 9

  BakalbaevaA. (2024). Problems of ensuring food security in the southern region of Kyrgyzstan. J. Osh State Univ. Econ.2, 81–88.

  • Google Scholar

• 10

  BektenovaD. C.DzhumakovaA. A. (2022). “Financial incentives for developing a green economy in the Kyrgyz Republic” in Sustainable development of the Republic of Kazakhstan under transition to green economy: Application of European Union countries experience, Eds. Makysh, S.B., Luniewski, S., Luniewski, A., Majdyrova, A.B., Nasyrova, G.A., Sembieva L.M., et al. (Astana: Eurasian National University named after L.N. Gumilyov).

  • Google Scholar

• 11

  BergA.BorenszteinE.SahayR.ZettelmeyerJ. (2024). “The evolution of output in transition economies: explaining the differences” in The economics of transition: The fifth Nobel symposium in economics, Eds. Berglöf, E., Roland, G. (London: Palgrave Macmillan UK).

  • Google Scholar

• 12

  BhatS. A.HuangN. F. (2021). Big data and AI revolution in precision agriculture: survey and challenges. IEEE Access9, 110209–110222. doi: 10.1109/ACCESS.2021.3102227

  • CrossRef
  • Google Scholar

• 13

  BKT. (2025) Digital agriculture and technology 4.0: Where are we really? Available online at: https://www.bkt-tires.com/ru/ru/blog/agriculture-blog/tsifrovoe-selskoe-khozyaystvo-i-tekhnologii-40-gde-my-na-samom-dele-nakhodimsya (Accessed March 12, 2025)

  • Google Scholar

• 14

  BodkerS. (2021). Through the interface: A human activity approach to user interface design. Boca Raton, FL, USA: CRC Press.

  • Google Scholar

• 15

  BogatyrevaA. S.YarkovaN. N. (2024). Yield programming and adaptive crop production. Perm, Russia: Permian: Perm State Agro-Technological University.

  • Google Scholar

• 16

  BondarI. L.LogachevM. S. (2025). Features of digital transformation of private farming. Russ. J. Resour. Conserv. Recycling12:01INOR325. doi: 10.15862/01inor325

  • CrossRef
  • Google Scholar

• 17

  BuedeD. M.MillerW. D. (2024). The engineering design of systems: Models and methods. Hoboken, New Jersey, USA: John Wiley & Sons.

  • Google Scholar

• 18

  BunkinA. F.PershinS. M.GudkovS. V.SarimovR. M.AstashevM. E. E.PishchalnikovR. Y. E.et al. (2024). Modern physical methods and technologies in agriculture. Physics-Uspekhi67, 194–210. doi: 10.3367/UFNe.2023.09.039577

  • CrossRef
  • Google Scholar

• 19

  ChernyshevaK. V.KorolkovaA. P.AfanasyevaS. I.KarpuzovaN. V. (2022). The use of data storage in the economy and management of the agribusiness. Mach. Equip. Rural Area4, 44–48. doi: 10.33267/2072-9642-2022-4-44-48

  • CrossRef
  • Google Scholar

• 20

  Climate FieldView. (2025) Maximize results with our digital farming solution. Available online at: https://climate.com/en-us/pricing.html (Accessed September 30, 2025)

  • Google Scholar

• 21

  CzvetkóT.KummerA.RuppertT.AbonyiJ. (2022). Data-driven business process management-based development of industry 4.0 solutions. CIRP J. Manuf. Sci. Technol.36, 117–132. doi: 10.1016/j.cirpj.2021.12.002

  • CrossRef
  • Google Scholar

• 22

  Da SilveiraF.LermenF. H.AmaralF. G. (2021). An overview of agriculture 4.0 development: systematic review of descriptions, technologies, barriers, advantages, and disadvantages. Comput. Electron. Agric.189:106405. doi: 10.1016/j.compag.2021.106405

  • CrossRef
  • Google Scholar

• 23

  DathanB.RamnathS. (2025). Object-oriented analysis, design and implementation. Cham, Switzerland: Springer Nature Switzerland.

  • Google Scholar

• 24

  Digital Tools. (2025). John Deere. Available online at: https://www.deere.com/en/digital-tools (Accessed September 30, 2025)

  • Google Scholar

• 25

  DJI. (2025). DJI Agriculture - Better Growth, Better Life. Available online at: https://ag.dji.com (Accessed September 30, 2025)

  • Google Scholar

• 26

  FortinoG.SavaglioC.SpezzanoG.ZhouM. (2020). Internet of things as system of systems: a review of methodologies, frameworks, platforms, and tools. IEEE Trans Syst Man Cybern Syst51, 223–236. doi: 10.1109/TSMC.2020.3042898

  • CrossRef
  • Google Scholar

• 27

  Fuentes-PeñaililloF.GutterK.VegaR.SilvaG. C. (2024). Transformative technologies in digital agriculture: leveraging internet of things, remote sensing, and artificial intelligence for smart crop management. J. Sens. Actuator Networks13:39. doi: 10.3390/jsan13040039

  • CrossRef
  • Google Scholar

• 28

  GebresenbetG.BosonaT.PattersonD.PerssonH.FischerB.MandalunizN.et al. (2023). A concept for application of integrated digital technologies to enhance future smart agricultural systems. Smart Agric. Technol.5:100255. doi: 10.1016/j.atech.2023.100255

  • CrossRef
  • Google Scholar

• 29

  GhoshP.KumpatlaS. (2022). “GIS applications in agriculture” in Geographic Information Systems and Applications in Coastal Studies. Eds. Y. Zhang and Q. Cheng. (London, UK: IntechOpen). 89–114

  • Google Scholar

• 30

  GrankinaV.VasilyevD. (2024). Digital transformation modeling for agricultural land irrigation. BIO Web Conf141:2005. doi: 10.1051/bioconf/202414102005

  • CrossRef
  • Google Scholar

• 31

  Guerrero-UlloaG.Rodríguez-DomínguezC.HornosM. J. (2023). Agile methodologies applied to the development of internet of things (IoT)-based systems: a review. Sensors23:790. doi: 10.3390/s23020790,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 32

  HaggagW. M. (2021). Agricultural digitalization and rural development in COVID-19 response plans: a review article. Int. J. Agric. Technol.17, 67–74.

  • Google Scholar

• 33

  ICT. (2025) National Statistical Committee of the Kyrgyz Republic. Available online at: https://stat.gov.kg/en/statistics/informacionno-kommunikacionnye-tehnologii (Accessed March 25, 2025)

  • Google Scholar

• 34

  Industry and Specialized 1C. (2025). Industry-specific and specialized 1C:Enterprise solutions. Available online at: https://solutions.1c.ru/catalog/erpapk (Accessed September 30, 2025)

  • Google Scholar

• 35

  KabakovV. V. (2022). Ensuring information security by building combined access control and management systems. Inform. Security Questions1, 7–10. doi: 10.52190/2073-2600_2022_1_7

  • CrossRef
  • Google Scholar

• 36

  KarbekovaA. B.KarbekovaM. J. (2020). Main indicators of the state of Crup production in the Kyrgyz Republic. Sci. Educ. Technol.2, 48–58.

  • Google Scholar

• 37

  KendyalaS. H.JoshiA.MallelaI. R.SinghS. P.JainS.GoelO. (2023). High availability strategies for identity access management systems in large enterprises. IJCSPUB13:5074869. doi: 10.2139/ssrn.5074869

  • CrossRef
  • Google Scholar

• 38

  KiryushinV. I.KiryushinS. V. (2023). Agricultural technologies. Saint Petersburg: Lan Publishing House.

  • Google Scholar

• 39

  KoshuevaK. B.ZhumabaevaT. T. (2023). Assessment of soil fertility status in greenhouses (on the example of the enterprise for municipal improvement and amenity planting of Osh City, Kyrgyz Republic). Bull. Sci. Pract.9, 168–174. doi: 10.33619/2414-2948/96/23

  • CrossRef
  • Google Scholar

• 40

  KotlyarovaE. G.StupakovA. G.ShiryaevA. V.KovalevaE. V. (2022). General agriculture, crop production. Belgorod: Belgorod State Agricultural University named after V. Y. Gorin.

  • Google Scholar

• 41

  KourV. P.AroraS. (2020). Recent developments of the internet of things in agriculture: a survey. IEEE Access8, 129924–129957. doi: 10.1109/ACCESS.2020.3009298

  • CrossRef
  • Google Scholar

• 42

  KozhomkulovaD. N.OganovaG. O.AdylbekovaN. O.AshymbaevaT. B.ShergazievaA. A. (2023). “Development of agricultural cooperation as the basis for digitalization of the agricultural sector of the Kyrgyz Republic” in Digital agriculture for food security and sustainable development of the Agro-industrial complex, Eds. Popkova, E.G., Sergi, B.S., Bogoviz, A.V., Semenova, I.E. (Cham: Springer). doi: 10.1007/978-3-031-27911-9_15

  • CrossRef
  • Google Scholar

• 43

  KujawaS.NiedbałaG. (2021). Artificial neural networks in agriculture. Agriculture11:497. doi: 10.3390/agriculture11060497

  • CrossRef
  • Google Scholar

• 44

  KumarA.SharmaM. J.MahajanA.KumarJ. (2024). Principles and practices for modern agronomy. Bhopal, India: Academic Guru Publishing House.

  • Google Scholar

• 45

  KyeremehK. (2021). Overview of system development life cycle models. J. Manag. Sci.11, 12–22. doi: 10.26524/jms.11.3

  • CrossRef
  • Google Scholar

• 46

  LinL.GuT.ShiY. (2024). The influence of new quality productive forces on high-quality agricultural development in China: mechanisms and empirical testing. Agriculture14:1022. doi: 10.3390/agriculture14071022

  • CrossRef
  • Google Scholar

• 47

  LogachevM. (2024). Functional requirements for an intelligent system for monitoring the impact of beaver activity on ecosystems. BIO Web Conf130:08022. doi: 10.1051/bioconf/202413008022

  • CrossRef
  • Google Scholar

• 48

  LogachevM.SimonovV. (2024). Information system as used to monitor sowing, care and harvesting on agricultural lands. BIO Web Conf83:03002. doi: 10.1051/bioconf/20248303002

  • CrossRef
  • Google Scholar

• 49

  MaatovaZ. M.OrozalievaA. A.TarykchievZ. M. (2025). Problems in the formation and utilization of labor resources in the agricultural sector of the Kyrgyz Republic. Econ. Bus. Theory Pract.6, 101–109. doi: 10.24412/2411-0450-2025-6-101-109

  • CrossRef
  • Google Scholar

• 50

  ManidaM.GaneshanM. K. (2021). New agriculture technology in modern farming. IJMRSS8, 109–114. doi: 10.30726/ijmrss/v8.i3.2021.83016

  • CrossRef
  • Google Scholar

• 51

  MatyukN. S.ShevchenkoV. A.MazirovM. A.PolinV. D.NikolaevV. A.Savos'kinaO. A. (2021). Adaptive agricultural technologies for cultivating field crops.Moscow: Russian State Agrarian University - Moscow Timiryazev Agricultural Academy (RSAU-MTAA).

  • Google Scholar

• 52

  MeshramV.PatilK.MeshramV.HanchateD.RamktekeS. D. (2021). Machine learning in agriculture domain: a state-of-art survey. Artif. Intel. Life Sci.1:100010. doi: 10.1016/j.ailsci.2021.100010

  • CrossRef
  • Google Scholar

• 53

  MirlanbekU.MaksudaK.AlimzhanI.GulbarA.SalmakhanS.ElviraS.et al. (2023). Some aspects of the government support problems in agricultural sector in the Kyrgyz economy. Int. J. Profes. Bus. Rev.8:69. doi: 10.26668/businessreview/2023.v8i5.1785

  • CrossRef
  • Google Scholar

• 54

  MowlaM. N.MowlaN.ShahA. S.RabieK. M.ShongweT. (2023). Internet of things and wireless sensor networks for smart agriculture applications: a survey. IEEE Access11, 145813–145852. doi: 10.1109/ACCESS.2023.3346299

  • CrossRef
  • Google Scholar

• 55

  MusarovaZ. A.AdamkulovaC. U. (2023). Geopolitics in the region: risks and prospects for ensuring food security in the Kyrgyz republic. Proc. Eng. Sci.5, 423–430. doi: 10.24874/PES.SI.02.018

  • CrossRef
  • Google Scholar

• 56

  NortonG. W.AlwangJ.MastersW. A. (2021). Economics of agricultural development: World food systems and resource use: New York, USA: Routledge.

  • Google Scholar

• 57

  NowakB. (2021). Precision agriculture: where do we stand? A review of the adoption of precision agriculture technologies on field crops farms in developed countries. Agric. Res.10, 515–522. doi: 10.1007/s40003-021-00539-x

  • CrossRef
  • Google Scholar

• 58

  OrlikL.KrasnikovaI. (2024). Exploring the functional possibilities of intelligent systems for active recreation. BIO Web Conf145:06011. doi: 10.1051/bioconf/202414506011

  • CrossRef
  • Google Scholar

• 59

  PandeyN.de ConinckH.SagarA. D. (2022). Beyond technology transfer: innovation cooperation to advance sustainable development in developing countries. WIREs Energy Environ.11:e422. doi: 10.1002/wene.422

  • CrossRef
  • Google Scholar

• 60

  PandeyP. C.PandeyM. (2023). Highlighting the role of agriculture and geospatial technology in food security and sustainable development goals. Sustain. Dev.31, 3175–3195. doi: 10.1002/sd.2600

  • CrossRef
  • Google Scholar

• 61

  ParpievaN.MatikeyevaN.SheralievaZ.AdylbekovaN.AmatovaU. (2023). Resource potential for the sustainable development of agriculture in the Kyrgyz Republic. E3S WoC380:01022. doi: 10.1051/e3sconf/202338001022

  • CrossRef
  • Google Scholar

• 62

  PasechnikovS. V.AvdeevV. V. (2020). Analysis of problems of granting rights access to information systems. System Administrator5, 93–95.

  • Google Scholar

• 63

  Projects. (2025) Ministry of Water Resources, agriculture and processing industry of the Kyrgyz Republic. Available online at: https://agro.gov.kg/industries/2 (Accessed March 1, 2025)

  • Google Scholar

• 64

  PryazhnikovaO. N. (2023). Digital agriculture in Africa: main areas of development. Econ. Soc. Probl. Russia1, 86–100. doi: 10.31249/espr/2023.01.04

  • CrossRef
  • Google Scholar

• 65

  QamarR.ZardariB. A. (2023). Artificial neural networks: an overview. Mesopotamian J. Computer Sci., 2023:124–133. doi: 10.58496/MJCSC/2023/015

  • CrossRef
  • Google Scholar

• 66

  RabhiL.JabirB.FalihN.AfraitesL.BouikhaleneB. (2025). Digital transformation metamodel in smart farming: crop classification prediction based on recurrent neural network. Foods Raw Mater.13, 107–118. doi: 10.21603/2308-4057-2025-1-626

  • CrossRef
  • Google Scholar

• 67

  RaihanA. (2024). A systematic review of geographic information systems (GIS) in agriculture for evidence-based decision making and sustainability. Global Sustain. Res.3, 1–24. doi: 10.56556/gssr.v3i1.636

  • CrossRef
  • Google Scholar

• 68

  RayhanaR.XiaoG.LiuZ. (2021). RFID sensing technologies for smart agriculture. IEEE Instrument. Measur. Magazine24, 50–60. doi: 10.1109/MIM.2021.9436094

  • CrossRef
  • Google Scholar

• 69

  Romero-GainzaE.StewartC. (2023). Ai-driven validation of digital agriculture models. Sensors23:1187. doi: 10.3390/s23031187,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 70

  SaparovaG.KhanG. D.JoshiN. P. (2024). Linking farmers to markets: assessing small-scale farmers' preferences for an official phytosanitary regime in the Kyrgyz Republic. Econ. Anal. Policy81, 696–708. doi: 10.1016/j.eap.2023.12.016

  • CrossRef
  • Google Scholar

• 71

  SemenovS. R.SemenovN. S. (2022). Development of information and consulting services in the system of information relations in the agricultural industry of the Kyrgyz Republic. Proc. Natl. Acad. Sci. Belarus Agrar. Ser.60, 23–34. doi: 10.29235/1817-7204-2021-60-1-23-34

  • CrossRef
  • Google Scholar

• 72

  SemenovS. R.SemenovN. S.AdievaA. A.TuleevR. D. (2025). Development of the export potential of the agricultural industry of the Kyrgyz Republic in the context of integration. AIP Conf. Proc.3276:20013. doi: 10.1063/5.0268242

  • CrossRef
  • Google Scholar

• 73

  Sencrop. (2025) Idroplan by farm technologies - DST for grapevine and fruit trees. Available online at: https://sencrop.com/eu/dst/idroplan (Accessed September 30, 2025)

  • Google Scholar

• 74

  ShaiyldaevaA. K.ChortombaevU. T.ShakirovaK. K.NazarbekovaE. U.KasymovaN. O. (2024). Economic analysis of permanent plants by categories of farms of the Kyrgyz Republic. BIO WoC83:07001. doi: 10.1051/bioconf/20248307001

  • CrossRef
  • Google Scholar

• 75

  SharmaK.ShivanduS. K. (2024). Integrating artificial intelligence and internet of things (IoT) for enhanced crop monitoring and management in precision agriculture. Sensors Int.5:100292. doi: 10.1016/j.sintl.2024.100292

  • CrossRef
  • Google Scholar

• 76

  ShcherbakovaA. S.IvanovV. A.MaltsevaI. S.OzarovskayaA. V. (2024). Digital transformation of agriculture: from global to regional level. J. Agric. Environ.11:21. doi: 10.60797/JAE.2024.51.10

  • CrossRef
  • Google Scholar

• 77

  ShpedtA. A.AksenovaY. V. (2021). Modern assessment of soil resources of Kyrgyzstan. IOP Conf. Series624:12233. doi: 10.1088/1755-1315/624/1/012233

  • CrossRef
  • Google Scholar

• 78

  SmirnovE. N.AntropovaM. Y. (2022). Scales and trends of digital transformation in global industry. Vestn. Univ.5, 53–60. doi: 10.26425/1816-4277-2022-5-53-60

  • CrossRef
  • Google Scholar

• 79

  SpielmanD.LecoutereE.MakhijaS.Van CampenhoutB. (2021). Information and communications technology (ICT) and agricultural extension in developing countries. Annu. Rev. Resour. Econ.13, 177–201. doi: 10.1146/annurev-resource-101520-080657

  • CrossRef
  • Google Scholar

• 80

  SridharA.BalakrishnanA.JacobM. M.SillanpääM.DayanandanN. (2023). Global impact of COVID-19 on agriculture: role of sustainable agriculture and digital farming. Environ. Sci. Pollut. Res.30, 42509–42525. doi: 10.1007/s11356-022-19358-w,

  • Pubmed Abstract
  • CrossRef
  • Google Scholar

• 81

  Syngenta Cropwise. (2025). Personalized approach with seed selector. Available online at: https://www.cropwise.com/russia (Accessed September 30, 2025)

  • Google Scholar

• 82

  TayeM. M. (2023). Theoretical understanding of convolutional neural network: concepts, architectures, applications, future directions. Computation11:52. doi: 10.3390/computation11030052

  • CrossRef
  • Google Scholar

• 83

  VolkovaI. N. (2021). Greenhouse industry of Russia and factors influencingits development and localization. Geograph. Environ. Living Syst.1, 93–109. doi: 10.18384/2712-7621-2021-1-93-109

  • CrossRef
  • Google Scholar

• 84

  YangH.JiangJ.ChenG.ZhaoJ. (2023). Dynamic load identification based on deep convolution neural network. Mech. Syst. Signal Process.185:109757. doi: 10.1016/j.ymssp.2022.109757

  • CrossRef
  • Google Scholar

• 85

  ZhangL.DabipiI. K.BrownJ. W. L. (2025). “Internet of things applications for agriculture” in Internet of things A to Z: Technologies and applications, Ed. Q. F. Hassan. IEEE: New York, USA. 13, 327–343. doi: 10.1002/9781394280490.ch13

  • CrossRef
  • Google Scholar

• 86

  ZhangX.ZhangW.NingZ. (2023). The evolving status of agricultural products in the global value chain since China and Kyrgyzstan established diplomatic relations. Sustainability15:4359. doi: 10.3390/su15054359

  • CrossRef
  • Google Scholar

• 87

  ZholbolduyevaD. S.AskarovaA. K.MurzalievaE. I.OmurzakovaU. M. (2024). Analysis of the current state and features of agriculture in the Kyrgyz Republic. Int. J. Human. Natural Sci.1, 94–98. doi: 10.24412/2500-1000-2024-1-4-94-98

  • CrossRef
  • Google Scholar

• 88

  ZhuravlevaL. A.ZarubinaE. V.ChupinaI. P.SimachkovaN. N.RuchkinA. V. (2022). Farm labor: characteristics, values, and problems. Moscow Econ. J.7:30. doi: 10.55186/2413046X_2022_7_3_156

  • CrossRef
  • Google Scholar