Research Article

A Systematic Literature Review on the Uses, Benefits, Challenges, and Prospects of Digital Twins in Livestock Farm Management

Md Kamrul Hasan1,2, Hong-Seok Mun1,3, Keiven Mark Bigtasin Ampode1,4, Eddiemar Baguio Lagua1,5, Hae-Rang Park1,5, Young-Hwa Kim6, Md Sharifuzzaman1,7, Chul-Ju Yang1,5*

1Animal Nutrition and Feed Science Laboratory, Department of Animal Science and Technology, Sunchon National University, Suncheon, Republic of Korea; 2Department of Poultry Science, Sylhet Agricultural University, Sylhet, Bangladesh; 3Department of Multimedia Engineering, Sunchon National University, Suncheon, Republic of Korea; 4Department of Animal Science, College of Agriculture, Sultan Kudarat State University, Tacurong, Philippines; 5Interdisciplinary Program in IT-Bio Convergence System (BK21 Plus), Sunchon National University, Suncheon, Republic of Korea; 6Interdisciplinary Program in IT-Bio Convergence System (BK21 Plus), Chonnam National University, Gwangju, Republic of Korea; 7Department of Animal Science and Veterinary Medicine, Bangabandhu Sheikh Mujibur Rahman Science and Technology University, Gopalganj, Bangladesh.

Md Kamrul Hasan and Hong-Seok Mun contributed equally to this work.

Received | May 26, 2024; Accepted | June 27, 2024; Published | October 23, 2024

*Correspondence | Chul-Ju Yang, Animal Nutrition and Feed Science Laboratory, Department of Animal Science and Technology, Sunchon National University, Suncheon, Republic of Korea; Email: yangcj@scnu.ac.kr

Citation | Hasan MK, Mun HS, Ampode KMB, Lagua EB, Park HR, Kim YH, Sharifuzzaman M, Yang CJ (2024). A systematic literature review on the uses, benefits, challenges, and prospects of digital twins in livestock farm management. Adv. Anim. Vet. Sci. 12(12): 2301-2314.

DOI | https://dx.doi.org/10.17582/journal.aavs/2024/12.12.2301.2314

ISSN (Online) | 2307-8316; ISSN (Print) | 2309-3331

Copyright: 2024 by the authors. Licensee ResearchersLinks Ltd, England, UK.

This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).

INTRODUCTION

According to the United Nations (World Population Prospects 2022: Summary of Results, 2022), it is assumed that the total population of the world will be 8.5, 9.7, and 10.4 billion in the year 2030, 2050, and 2100 respectively. It has been predicted that in order to meet the projected food demand resulting from population growth, global crop production will need to at least double by 2050 (Tilman et al., 2011; Alexandratos and Bruinsma, 2012). Different agricultural technologies with the combination of Information and Communication Technology (ICT) are developing to provide sufficient food for the ever-increasing population (Saidu et al., 2017; Mat et al., 2018; Farooq et al., 2019; Farooq et al., 2022). Agriculture 4.0 (da Silveira et al., 2021) is a part of industrial and agricultural revolutions and driven by technological advancements such as edge computing (Zhang et al., 2022), artificial intelligence (AI) (Smith, 2018), real-time sensing and wireless sensor networks (WSNs) (Loukatos and Arvanitis, 2021), and the internet of things (IoT) (Paraforos and Griepentrog, 2021), which are opening the door for data-driven farming. With its emphasis on sustainable development, Agriculture 5.0 (Kaklauskas, 2018) is generating interest in digital twin (DT) and their potential applications in agriculture (Tzachor et al., 2022). DT is the digital and virtual counterparts of any physical object (Fuller et al., 2020). Nowadays it is used in different industries (Errandonea et al., 2020), and can track and forecast the state and condition of the machines (Aivaliotis et al., 2019). DT is used to monitor animal behavior (Zhang et al., 2023) with future predictions (Han et al., 2022), reduce methane emissions from dairy farms (An and Chen, 2021), optimize production lines in poultry feed factories (Tarek et al., 2022), and manage feeding (Raba et al., 2022) and environment control (Jo et al., 2019; Jeong et al., 2023) in animal farms. Various ICTs, IoT, cameras, radiofrequency identification (RFID), and sensors are used in precision livestock farming (PLF) to identify animals and behavior, monitor farm animals and the environment, detect diseases early, and manage farms optimally (Besteiro et al., 2018; Pezzuolo et al., 2018; Benjamin and Yik, 2019; da Fonseca et al., 2020; Hristov et al., 2021; Pandey et al., 2021; Haldar et al., 2022).

Several review articles about the use of DT in agriculture have been published (Sreedevi and Santosh Kumar, 2020; Pylianidis et al., 2021; Verdouw et al., 2021; Nasirahmadi and Hensel, 2022; Peladarinos et al., 2023). To the best of the authors’ knowledge, there is not enough research that addresses the use of DT in livestock farm management specifically. This study aims to evaluate research progress and application of DT in livestock farm management by analyzing recent case studies and to guide future investigations into the advancement of DT.

MATERIALS AND METHODS

A systematic review was carried out in compliance with PRISMA guidelines (Page et al., 2021), and by using the methodology suggested by systematic and bibliometric reviews (Agnusdei et al., 2021; Salazar-Moya and Garcia, 2021; Ahn and Kim, 2023). The review was carried out in five stages: firstly, identifying the research questions; secondly, choosing sources of information; thirdly, organizing the search process; fourthly, specifying the qualifying requirements; and finally, choosing the papers.

Identifying the Research Questions

The primary aim of this paper is to conduct a literature search on the use of DT in livestock farm management and to evaluate the research findings in order to identify the prospects and obstacles associated with this area of study. The following research questions (RQs) therefore serve as a summary of these objectives.

RQ1. What are the advantages of deploying DT for animal behavior monitoring and environment control?

RQ2. What are the economic, social, technological, ethical, and environmental challenges of deploying DT in livestock farms?

RQ3. What purposes does the livestock farm management hope to achieve from DT?

Choosing Sources of Information

Three academic research databases were searched: Google Scholar, Web of Science (WoS), and Scopus. They are well known for maintaining strict selection criteria, only accepting scholarly content from high-quality journals and conference proceedings, thereby ensuring the inclusion of top-tier research. Indirect inclusion was made for contributions that are indexed in ScienceDirect, IEEE-Xplore, and SpringerLink.

Organizing the Search Process

Various keywords closely aligned with the research inquiries were utilized during the literature search process. The phrase “digital twin” was chosen as the main keyword. There were other keywords which are as follows: “livestock farming”, “application”, “cattle”, “pig”, “welfare”, and “environment”. These keywords were linked to the Boolean operators “AND” for better searching of literature. There was keyword selection bias because synonymous keywords were not used across different databases. Moreover, there were language and time frame limitations in this study due to its focus exclusively on English-language articles published within the last five years.

Specifying the Qualifying Requirements

Articles from peer-reviewed journals, conference proceedings, and doctoral consortiums that highlight the uses of DT in the livestock sector were included. Reviews of the literature and comparative studies were not included because this study focused on trial-based results. The literature search was restricted to English-language works, and contributions from all over the world in the last 5 years (2018-2023) were included due to the recent development of DT in livestock farm management. Conference proceedings as well as accepted and published articles from journals with an index were included.

 

Choosing the Papers

A total of 393 papers were produced because of the initial selection of contributions based only on the topic of “digital twin” and “livestock farming” with a time frame from 2018 to 2023; of these, 360 were indexed in Google Scholar, 15 in the WoS, and 18 in the Scopus. The use of DT in livestock farm management was our topic and considering the research questions a total of 325 papers were eliminated from the Google Scholar searches based on searches produced by connecting the specified keywords- “livestock farming”, “application”, “cattle”, “pig”, “welfare”, and “environment”. A total of 68 papers were selected (35 papers were Google Scholar-indexed, 15 papers were WoS-indexed, and 18 papers were Scopus-indexed). Papers containing duplicate submissions, titles exhibiting a close resemblance, and content failing to meet the primary criteria were subsequently excluded through the meticulous screening of their titles, abstracts, and keywords. Manual examination was used to find and remove duplicate entries from each article. After a thorough review of the title and abstracts, the publication details were examined in order of priority. Due to the use of three databases in this investigation, duplicate articles were found that appeared repeatedly in different databases. Prioritizing research questions led to the establishment of distinct inclusion and exclusion criteria, which were tailored to specific topics, experimental designs, publication types, and involved stages such as title and abstract review, full-text assessment, real-world application, and the initial phases of DT model development. As a result, the relevant article came down to 39. A methodical process of removing irrelevant articles was followed, with an emphasis on including those that showed practical applicability in actual farming environments. Articles with potential for the future but not immediately applicable to farm conditions today were also included in the study. To ascertain the relevance and accomplishment of the study’s objectives and to reduce the chance of missing any crucial information, each of these 39 papers was examined separately. A total of 13 papers were kept for this study, after being carefully evaluated for their scope, the particular subject matter, and the required qualifications. The procedure for choosing papers is schematically depicted in line with the PRISMA flowchart (Figure 1).

Use of Digital Twin in Livestock Farm Management

According to the analysis of the 13 papers published recently on the use of DT in livestock farm management, three of the papers concentrated primarily on animal behavior and welfare monitoring, three on real-time environmental and energy control in livestock farms, and three on livestock supply chains. Furthermore, there were two papers addressing livestock farm operation, one focusing on decision-making processes concerning animal feed factories, and another centered-on strategies for mitigating greenhouse gas emissions from livestock farms. Different physical objects were used in various papers, with cattle appearing as more specific animal species of interest. All these papers—aside from (Keates, 2019) and (Tarek et al., 2022)—have a close connection with animal welfare and management.

To identify the feeding behavior of cows, Zhang et al. (2023) suggested a pioneering architectural framework for cows utilizing DT technology, designed to encapsulate the entirety of the cow’s lifecycle. A method was used to create the cows’ digital shadow by gathering real-time location and neck movement data from cows of different stages (calf, heifer, and adult) using an inventive custom-made intelligent collar that included ultra-wideband (UWB) chips and inertial measurement units (IMUs). Based on the data dependency, the whole architecture was classified into three levels: barn integration, pasture agent, and cloud. Besides, there were five layers- physical entity, integrated, data management and information, modeling and simulation, and decision and user interface. They used three machine learning (ML) models- support vector machines (SVM), k-nearest neighbor (KNN), and a deep learning (DL) model- long short-term memory (LSTM), and the accuracy was highest in LSTM (94.97%). Petrov and Atanasova (2021), created a 3D real-time virtual reality (VR)/ augmented reality (AR) model of farm animals to monitor animal behavior and welfare. Following the five domains for assessing animal welfare—the physical environment, health, mental state, behavioral interactions, and nutrition—the model was created using data gathered from IoT devices. ML software was used to analyze the data in a meaningful way, alerting farmers when values seemed to deviate from the predetermined ranges. The software also offers details on potential issues that could result from the measured deviations as well as prompt action items for resolving the issue. At its nascent developmental phase, this model has yet to showcase any evaluation outcomes. Han et al. (2022) implemented DT for the identification and future prediction of different behaviors of cattle- resting, rumination, high activity, medium activity, panting, eating, grazing, and walking. Through the utilization of sensor data retrieved from an IoT farm system, this innovative solution constructs a DT of cattle utilizing DL (LSTM), enabling real-time monitoring of the physiological cycles of cattle, and forecasting the state of subsequent physiological cycles. This model exhibited a training error of 0.58 with a future prediction error of 5.12, and their findings indicate that DT can predict the future behavior of cattle, which can help the farmers in monitoring animal health and welfare.

Jo et al. (2018) conducted a feasibility study for controlling the environment of the pigsty by using DT technology. The two-tiered framework of smart livestock farms—the digital farm engine and the digital farm framework—was outlined, and a brief description of an easily deployable service scenario for automatic control in livestock farms was given to illustrate how physical and digital livestock farms interact. They define a DT for pig enlargement optimization, which was achieved by controlling a barn system that maintains temperature and air quality within a pre-set collection. To find scenarios that lead to the desired outcome, a combination of big data and model-based simulations were used to provide decision support and mechanize the pigsty control system. The farm control mechanism was needed to control both automatic window openings and airflow to achieve the desired state. They found that DT can efficiently regulate the animal shed’s environment, helping to improve the welfare of animals on livestock farms. The DT of this framework is in the initial stage of development, and till now it is not applied in real livestock farm conditions. Moreover, Jo et al. (2019) proposed a DT to simulate a pigsty’s energy consumption to offer decision support for the best pigsty construction with optimal energy use. Six different models of ventilation fans (SLF-300D2-6, SLF-350A4-6, SLF-500A4-6, SLF-500D4-6, SLF-730A6-5, and SLF-960A6-3) were utilized in the pigsty. EnergyPlus, an energy simulation program used to model energy consumption for heating, cooling, and ventilation, was employed in the pigsty. A variety of ventilation fan models were employed in combination to determine the optimal configuration for energy savings. Among the different fan combinations, case number 4 (2 fans of the SLF-300D2-6 model and 4 fans of the SLF-960A6-3 model) exhibited the lowest electricity consumption (approximately 550 MJ). This study suggests that DT can reduce farm expenses by reducing electricity costs. Furthermore, Jeong et al. (2023) deployed DT in pig farms for making virtual pig houses which provided information for the efficient use of heating, ventilation, and air conditioning (HVAC) systems in the real pig house to maintain the best environment for the pigs, and helped to solve the problem related to HVAC system before it appears in the real pig house. Energy consumption of heating systems and ventilation fans in proportional, conventional, optimal, and fixed (60°C) operating methods were examined. DT technology aided in finding the best operating method which was the ‘optimal method’ that consumed the lowest amount of electricity for operating ventilation fans (67.01 kWh) and heating purposes (1051 kWh) than others. This study suggests that DT can effectively manage environmental conditions in animal sheds while minimizing electricity costs.

Mu et al. (2022) illustrated the results of a case study on the use of DT in China’s Qinghai Meadow, and the purpose of this case study was to create a livestock supply chain. The two essential elements that comprised the foundation of this approach were the commercial supply chain and the information technology (IT) operating system. The IT operating system underwent categorization into three distinct tiers: physical, data, and virtual. Information obtained from sensors within the physical layer underwent transmission to the data layer for subsequent processing, evaluation, management, and archival purposes. Subsequently, data processed within the data layer was relayed to the virtual layer, where advanced AI and ML models were employed for further data refinement, and then provided sophisticated feedback to the physical layer. The main purpose of this process was to help in early disease diagnosis and prevent losses from severe weather. For this reason, regardless of the animal or the surroundings, alarms were set off whenever the system noticed unusual behavior; however, this solution has not yet shown any evaluation outcomes. Besides, Keates (2019), described a DT to enhance livestock value chains, and the goal of this DT was to serve as a tool for gathering and disseminating important metrics and assessing each one’s performance in comparison to a reference model. One of the main use cases mentioned was the creation of missing data by simulating the actual supply chain using the reference model, which improved supply chain comprehension. The life cycle of meat animals, ranging from 2 to 10 years, posed significant challenges in collecting individual animal data. Due to the lack of individual animal data, this DT framework is still in its initial stages of development and has not yet been applied to actual circumstances. Furthermore, to attain ‘zero-waste,’ Valero et al. (2023), provided a framework of DT in several areas of a circular meat supply chain, one of which is animal management. The framework of DT in the animal management area can monitor, predict, and influence animal behavior and environmental conditions which provide valuable insights into the health and welfare of animals. The physical objects include information about individual animals (for example their activity, resting status, and grazing behavior) as well as environmental information in their shed (for example temperature, humidity, and lighting); however, this framework of DT is still at the conceptual level.

Erdélyi and Jánosi (2019), proposed a DT of the porker model in MATLAB Simulink environment, which can predict future feed intake, body weight gain, and final body weight of pigs. Details of the porker model were not explained in their paper. This DT framework can help farmers in their pig farm management and future planning; however, the DT of the porker model is still being developed, and it has not been used on a real pig farm yet. Raba et al. (2022) applied DT technology for better feed management and optimization of the animal feed delivery chain. This was accomplished at the farm level using a DT-based strategy, which remotely monitored the amount of feed in feed bins, using RGB-D (red, green, blue-depth) camera images, processed by combining biased-randomization techniques with a simheuristic framework. The results exhibited that both farmers and feed producers can remotely manage their feedstock levels using DT. This study suggests that by deploying DT on their farms, farmers can enhance feeding management. Tarek et al. (2022) implemented DT in a poultry feed manufacturing factory by using a four-layer software architecture (physical, data, analytical, and presentation layer) to monitor the whole feed manufacturing process. All the assets, for example, IoT sensors, machines, storage facilities, and containers, were included in the physical layer, whereas model definitions and the relationships between assets were contained in the data layer. The analytic layer was affixed to the data layer, and it offered service modules and provided a description of the feed factory’s assets and production line processes. The simulation model to replicate various factory operating conditions was included in the analytic layer, and the dashboard, which was used to display and examine data, was part of the presentation layer. Moreover, the graphs that exhibited the connections between the factory’s assets for poultry feed were included in the presentation layer. This DT technology helped in decision-making on whether to use single or multiple production lines to meet the demand for the quantity of feed production.

An and Chen (2021), used swarming unmanned aircraft vehicles (a group of drones) on the dairy farm to provide real-time data updates for the DT which assisted in the measurement and reduction of methane emissions from the dairy farm. To deploy the DT platform, they used web app, gazebo, robot operating system, and unmanned aerial vehicle toolbox. In addition, unmanned aerial vehicles were also utilized as mobile sensors and actuators to monitor the diffusion of methane and to minimize methane emissions by spraying methane neutralizers (biochar). At various altitudes, they observed the 3D flight path of an unmanned aerial vehicle waypoint within an unmanned aerial vehicle toolbox. Additionally, the entirety of the unmanned aircraft vehicles’ flight paths within the gazebo was observed. Finally, observations were made of the methane diffusion process. Their results suggest that greenhouse gas emissions from livestock farms can be reduced by using DT technology.

A summary of research on the application of DT in livestock farm management is shown in Table 1.

Benefits of using Digital Twin in Livestock Farm Management

Because only seven of the thirteen studies mentioned above have used DT technology in actual farm settings, the literature review exhibits that the technology is still in its early stages of development. However, DT technology has several benefits for livestock farm management. DT exhibited positive results in the animal shed model in operating ventilation fans and windows (Jo et al., 2018), therefore, farmers can use a laptop or smartphone to automatically control the farm environment from a distance.

According to Jo et al. (2019), the most energy-efficient ventilation fan combinations can be effectively detected using DT, and farmers can use these fan combinations in farms to reduce production costs. DT can be effectively harnessed to ascertain the optimal energy-efficient operating methods for running the HVAC system in animal sheds (Jeong et al., 2023) and through the identification and removal of operational inefficiencies, DT can assist farmers in reducing their farm expenses.

DT technology can monitor a comprehensive range of animal behaviors—including resting, rumination, high and medium activity levels, panting, eating, grazing, and walking—and can also provide predictive analytics (Han et al., 2022), thereby assisting farmers in enhancing overall farm management practices. With an accuracy of 94.97%, DT with LSTM model identified animals’ feeding behavior (Zhang et al., 2023), assisting farmers by providing important information about animal health and welfare status. Continuous monitoring of daily feed consumption can be successfully done by DT (Raba et al., 2022) and such kind of monitoring can assist farmers in early detection of animal health and welfare issues and making better future planning for farm operations. DT can be significantly used in making decisions regarding the selection of the production line (single or multiple) for manufacturing livestock feed (Tarek et al., 2022) which can assist farmers in manufacturing desired amounts of feed within a short period. DT can be successfully utilized to accurately measure and

 

Table 1: An overview of research on the application of digital twins in livestock farm management.

DT with AI model	Purpose	Field appli-cation	Physical object (No.)	Findings							Ref.
DT with ML (SVM, KNN), and DL (LSTM)	Identification of feeding and non-feeding behaviors.	Yes	Cattle (5)	Model	Accu-racy (%)		Recall (%)	Precision (%)	F1 score (%)	Specificity (%)	(Zhang et al., 2023)
				SVM	88.83		89	89.90	89.45	88.64
				KNN	90.53		90.20	92	91.09	90.91
				LSTM	94.97		93.86	99.99	95.21	99.99
DT with ML	To monitor animal behavior and welfare.	No	All livestock animal	DT assists in monitoring animal behavior and welfare.							(Petrov and Atanasova, 2021)
DT with ML	To build a livestock supply chain.	No	Cattle	Both early disease detection and prevention of catastrophic weather are possible by DT.							(Mu et al., 2022)
DT with DL (LSTM)	Identification and future prediction of behaviors.	Yes	Cattle (98)	Training error 0.58 Future prediction error 5.12							(Han et al., 2022)
DT of the livestock value chain	To create a DT to enhance Australia's livestock value chains.	No	Value chain of livestock	DT aids in gaining a deeper comprehension of the livestock supply chains.							(Keates, 2019)
DT of meat supply chain	To build a meat supply chain that has zero-waste.	No	Supply chain of meat	DT aids in reducing waste in the meat supply chain.							(Valero et al., 2023)
DT of the porker model	To build a DT for smart pig farming.	No	Pig	DT assists in the future prediction of feed consumption, body weight gain, and final body weight of fattening pigs.							(Erdélyi and Jánosi, 2019)
DT of dairy farm	To assist in quantifying and reducing the methane emissions from dairy farms.	Yes	Dairy farm (1)	DT can detect and reduce methane emissions.							(An and Chen, 2021)
DT of poultry feed factory	For smart manu-facturing of poultry feed.	Yes	Poultry feed factory (1)	DT assists in deciding the production line (single or multiple) to produce the required quantity of poultry feed.							(Tarek et al., 2022)
DT with sim-heuristic frame-work based on biased rando-mization tech-niques	To supply precise real time data for tracking daily feed con-sumption, and future planning to prevent runouts of feedstock. To alter the feed company's approach by shifting the maintenance of the feedstock's workload.	Yes	Feed bin (325)	Farmers can do better feeding management in livestock farms by using DT. DT assists feed producers in making better plans for future feed production.							(Raba et al., 2022)
DT of pigsty	To auto-matically regulate the pigsty's environment.	No	Pigsty	Decision-making by farmers regarding automated environmental control systems in pigsty is aided by DT.							(Jo et al., 2018)
DT of pigsty	Energy planning for automated environ-mental control in the pigsty.	Yes	Pigsty (1)	Case Number					Energy consumption (MJ)		(Jo et al., 2019)
				Case number 1 (10 fans of SLF-300D2-6 and 15 fans of SLF-350A4-6)					Approximately 710
				Case number 2 (6 fans of SLF-500A4-6 and 4 fans of SLF-500D4-6)					Approximately 850
				Case number 3 (4 fans of SLF-500D4-6 and 4 fans of SLF-730A6-5)					Approximately 750
				Case number 4 (2 fans of SLF-300D2-6 model and 4 fans of SLF-960A6-3 model)					Approximately 550
DT of pigsty	To run the HVAC system in a pigsty with minimal energy consumption.	Yes	Pigsty (1)	Operating methods		Electricity consumption of ventilation fans (kWh)			Heating energy consumption (kWh)		(Jeong et al., 2023)
				Proportional		95.41			1138.61
				Conventional		91.55			-
				Optimal		67.01			1051
				Fixed (60oC)		-			1288.61

 

SVM-support vector machines, KNN-k nearest neighbor, LSTM-long short term memory, HVAC-heating, ventilation, and air conditioning, MJ-Megajoule, and kWh-kilowatt-hour.

 

 

subsequently reduce methane emissions from animal farms (An and Chen, 2021), and assist farmers in solving environmental problems and creating more environmentally friendly farming practices that will lessen greenhouse gas emissions. The overall benefits of DT in different aspects of livestock farm management including farm operation, environmental control, and welfare monitoring are summarized in Figure 2.

Digital Twin Deployment Challenges and Possible Solutions

With the deployment of DTs, a variety of economic, social, technological, ethical, and environmental challenges have emerged (Neethirajan, 2023). In terms of economic challenges, high switching costs will be required to incorporate DT in livestock farms for the setup of sensors, devices, hardware, and software (Neethirajan and Kemp, 2021). The infrastructure for PLF varies, which makes it difficult to use DT in this setting (Mallinger et al., 2022). Round-the-clock internet facilities, the cost of laptops and desktop computers, training costs of farmers, additional energy costs, and other costs to run this technology effectively are to be added. There are deployment and operational costs to run the DT effectively which can vary based on farm size and the use of AI technologies. The deployment costs include PLF tools for monitoring farm animals and the environment, the development of AI software, farm structure development for setting PLF tools, and internet connectivity. Operational costs include annual maintenance fees (for updating software, and changes of broken sensors and devices), costs related to data (for integration, processing, and analysis), and training costs (for skill development and proper operation of DT). Government subsidies, credit options, and tax exemptions could aid farmers in overcoming their financial obstacles. For example, subsidies and tax incentives assisted in the quick adoption of different DT technologies in the manufacturing sector of Germany (Platform Industrie 4.0, 2024). It also helped to reduce the production cost and increase efficiency.

In the social aspect of challenges, if the DT is only accessible by large-scale farmers, then medium and small-scale farmers will be unable to establish it on their farms which may create a monopolization market, and there will be a high chance of increasing international dependencies (Mallinger et al., 2022; Malik et al., 2023). Moreover, it may create social discrimination among farmers which may affect the variation in the health and welfare status of animals (An and Chen, 2021). Farmers’ support to overcome social challenges could come from the domestic production of digital devices and technology, coupled with the provision of accessible credit facilities.

In terms of technological aspects of challenges, technical proficiency in fields of data analytics, IoT, and AI is required for operating DT effectively, and many farmers might not possess the required abilities (Mallinger et al., 2022); lacking assistance, and training. Appropriate installation of animal-attached sensors, for example, accelerometer, pedometer, radar-based sensor, and respiratory rate sensor is challenging (Neethirajan, 2020), and if they are not installed correctly, they may fall out of the body. Furthermore, most of the sensors rely on batteries for operation (Dewan et al., 2014); therefore, if the batteries run out of power (Oudenhoven et al., 2012), the sensor’s ability to provide real-time animal data may be compromised, potentially disrupting the entire DT system. ML and/or DL models are used in DT technology (Petrov and Atanasova, 2021; Han et al., 2022; Mu et al., 2022; Zhang et al., 2023), and these models rely on large and high-quality data (Zhang et al., 2018; Upadhyay and Khandelwal, 2019; Prior et al., 2020; Zhang et al., 2022); integrating these data is still challenging (Davis et al., 2017; Zhang et al., 2018; Jarrahi et al., 2023). When there is insufficient data, the DL model exhibits unsatisfactory accuracy due to its sensitivity to data sources (Han et al., 2022). Furthermore, improper labeling and the use of skewed training datasets can affect future animal behavior prediction by DT technology (Han et al., 2022). By implementing comprehensive training, creating a user-friendly interface, and optimizing and integrating AI models, farmers can effectively navigate and overcome technological hurdles.

In ethical aspects of challenges, special concern is needed about human-animal relationships (Neethirajan, 2023). The relationship between the farmer and the animal could suffer if DT is used because sensors and/or devices will gather all the data required to run the technology. DT technology uses the ML model which assesses different aspects of health and nutrition effectively, but it lacks the intuitive understanding of animal needs that humans have (Seibel et al., 2020). In PLF, PLF tools (sensors and devices) treat animals like objects and only gather certain parameters from them (Bos et al., 2018). ‘Treating as an object’ and ‘turning into an object’ are the two categories of objectification (Brom, 1997), and objectification using PLF tools has detrimental effects (Bos et al., 2018). ‘Treating as an object’ refers to treating an animal like a mere object and never considering its inherent worth (Bos et al., 2018). ‘Turning into an object’ refers to focusing more on the economic worth of animals (milk, meat, and eggs) and never considering their welfare (Bos et al., 2018). Consequently, the overall sustainable development of industrial farming can be impacted by viewing an animal as nothing more than an object and concentrating only on its financial worth. It is crucial to keep in mind that animals are sentient beings with social and emotional needs (Proctor, 2012; Proctor et al., 2013; Kumar et al., 2019), and using DT should never lead to a disrespect for animal welfare or a failure to uphold human responsibility for animals. The development of a DT model that prioritizes the social and emotional needs of animals may have the potential to address ethical challenges. Moreover, animal welfare guidelines for the ethical use of PLF tools should take into account the behavioral needs and economic characteristics of animals in order to achieve a balanced approach between PLF tools and animal ethics.

In terms of environmental challenges, it is still up for debate how best to evaluate and lessen the environmental impact of DT (Neethirajan, 2023). Along with carbon emissions from the manufacture and disposal of digital infrastructure and devices, the use of DT in livestock farming can also produce electronic waste (Liu et al., 2019; Neethirajan, 2023). Energy is needed for data processing and to run the DT even though this technology can assist with sustainability by automatic environment controlling of animal sheds with minimal energy consumption (Jo et al., 2019; Jeong et al., 2023). Maximizing the utility of digital devices alongside the exploration of alternative energy sources may present a viable strategy for farmers to surmount environmental challenges. For long-term sustainability, the use of renewable energy sources (solar, wind, and other new energy sources) for operating DT in livestock farms can minimize the environmental footprint of DT. For example, photo-voltaic power generation saved 2 × 104 KWh of electricity and 20 t of CO2 emissions annually from a dairy farm (Zhang et al., 2017). By integrating renewable energy sources and efficiency-improving methods, such as borehole thermal energy storage and heat pump technology, the photovoltaic-thermal system supplies the pig farm’s reliance on fossil fuels for energy, culminating in a decrease of 20,850 kg in CO2 emissions annually (Murali et al., 2024).

Moreover, there exist additional challenges. DT is used in monitoring animal behavior and welfare (Petrov and Atanasova, 2021; Han et al., 2022; Zhang et al., 2023), and environmental control (Jo et al., 2018; Jo et al., 2019; Jeong et al., 2023); however, it may be difficult to integrate all aspects of the livestock production cycle, spanning from animal breeding to product processing. In developing countries around 50% of farmers are still out of internet facilities (Arulmozhi et al., 2021), which is a major concern for establishing DT in livestock farms; furthermore, there are electricity problems (Roubík and Mazancová, 2020; Saha et al., 2024). Clear guidelines and regulations are necessary to address the complex issue of DT-generated data ownership (farmer and/ or technology provider) (Mallinger et al., 2022), and safe-guarding the privacy of sensitive information, for example, animal disease records and animal welfare status. Livestock farming is more challenging than manufacturing industries such as garments and shoe companies because all operations related to cloths and shoe production occur systemically in a controlled environment, whereas several risk factors such as natural calamities (flood, droughts), outbreaks of diseases, and changes in the consumption patterns of livestock products (milk, meat, and egg) can create a variation in the market demand (Neethirajan and Kemp, 2021). Therefore, farmers will deploy DT technology if this technology can reduce the effect of these unknown risk factors (Neethirajan and Kemp, 2021), which is very challenging. Farmers are confused about the validity and accuracy of DT and scientific evidence is scarce on how this technology will improve production performance with quality, and finally benefit. Facilitating farmers in overcoming additional challenges may involve collaborative research efforts, conducting further case studies, optimizing resource utilization, exploring alternative energy sources, ensuring robust cybersecurity, making policies for data ownership and usage, and harnessing the potential of blockchain technology. DT deployment challenges and possible solutions are shown in Figure 3.

Prospects of Digital Twin in Livestock Farm Management

Innovations in IoT, AI, cloud, and edge computing have the potential to assist farmers in real-time data processing. Data privacy and cybersecurity will increase the faith of farmers. Data integration and visualization can help farmers for better farm operation and planning. Additionally, the use of renewable energy sources for operating DT technology may minimize production costs and greenhouse gas emissions. User-friendly interfaces and regulatory standards will ensure the optimal use of DT while considering the animal welfare issue. These innovations are believed to drive the future of DT application in different aspects of livestock farming. Various AI models, for example, convolutional neural networks (CNN), ML, DL, artificial neural networks (ANN), adaptive neural fuzzy inference systems, and pattern recognition, are employed in precision livestock farming to identify animals and diseases, monitor farm animal behavior and welfare, observe daily body weight growth, and monitor the farm environment (White et al., 2004; Kashiha et al., 2014; Kashiha et al., 2014; Khoramshahi et al., 2014; Lee et al., 2019; Nasirahmadi et al., 2019; Alameer et al., 2020; García et al., 2020; Bao and Xie, 2022).Every single AI model has some limitations for the explanation and accuracy of results (Birhane, 2021), for example, the ML model exhibits biases (Alelyani, 2021; Chakraborty et al., 2021). Therefore, priority should be given in making an appropriate design that has no hidden biases that, through computational feedback mechanisms bination of two or more AI models in the design of DT technology, for example, deep hybrid learning (a combination of DL and ML), which may increase accuracy. could magnify unfavorable effects on livestock productivity

 

(Malik et al., 2023). In the future, there may be a comThe hybrid blended deep learning (HyBDL) model exhibited higher accuracy (98.03%) than CNN, gated recurrent units (GRU), LSTM, CNN-GRU, and CNN-LSTM for milk adulteration detection (Mhapsekar et al., 2024). Optimization of the AI model is essential for a higher level of accuracy (Abdolrasol et al., 2021). When the inertial measurement units (IMUs) and indoor position detection data were combined in the DL model of a DT technology, it identified the feeding behavior of cows with a higher degree of accuracy (94.97%) than when the IMUs data was used alone (91.05%) (Zhang et al., 2023), showing a considerable amount of improvement can be achieved. In the days ahead, anticipate the increasing sophistication and advancement of AI models that analyze different parameters related to animal health and welfare. Farm animal affective states and behaviors may be simulated and predicted in the future through the use of DT (Neethirajan, 2022). DT might be developed that could tailor care plans for specific animals according to their distinct behaviors, emotions, and health requirements. In the coming days, there will be species-specific biometric sensors (Neethirajan, 2022), and they will collect different parameters related to the health, welfare, production, and reproduction of animals that are needed to run the DT technology effectively. A next-generation method for achieving real-time flow of biometric data is offered by DT models of individual animals (Jo et al., 2018; Neethirajan and Kemp, 2021).

A DT reference model is necessary for the development of animal twins. Generally, the DT reference model framework consists of the physical object, communication layer, and digital object layer (Lu et al., 2020). The communication layer transfers data between the physical object and the digital object layer (Lu et al., 2020). The data processing section of the digital object layer processes the data that has been collected from the physical object by the information model of the digital object layer (Lu et al., 2020). For developing the DT reference model there are several steps: firstly, setting of the objectives and scope. Secondly, the identification of stakeholders and understanding of their expectations regarding the use of DT. Thirdly, strategy planning for data collection and integration using PLF tools. Fourthly, designing the architectural framework for data processing and storage. Fifthly, the selection of the appropriate ML model and visualization tools (mobile apps) for data analysis and user decision implementation. Sixth, accuracy and reliability testing for improving performance. Seventh, trail basis and farm condition testing to see applicability in the real farms. Training farmers and gathering feedback for efficient use and continuous improvement. Eighth, a user manual for written documentation and making maintenance schedules for better operation of the DT. There are no DT reference models that can direct research toward the creation of animal twins (Neethirajan, 2022). In the forthcoming days, researchers may create animal twins with the aid of DT reference models. Low latency and quick analysis are expected to be the future of DT using edge computing. Moreover, by protecting livestock farm data privacy, blockchain technology may help to increase faith among DT technology provider companies and farmers. Technology providers are expected to develop more user-friendly DT. Currently, most publications exploring the implementation of DT technology into livestock farms are in their embryonic phases, and in the foreseeable future, these foundational frameworks hold potential for application in livestock farming to enhance the credibility of this technology. The potential for efficient farm operations and improving animal health and welfare through DT technology is promising. This can be done by addressing the drawbacks and issues, looking into potential future directions, and making use of the automation and comprehensive management capabilities of DT.

According to predictions made by industry research firm Gartner (Top 10 Strategic Technology Trends for 2019: Digital Twins, 2024), DT technology was among the top ten strategic technology trends in 2019. The market value of DT is projected to reach 131.09 billion US dollars by 2030, which was only 6.3 billion US dollars in 2021 (Digital Twin Market, 2022). This indicates that in the future there will be more applications of DT technology in different industries including the livestock industry.

The potential for efficient farm operations and improving animal health and welfare through DT technology is promising. This can be done by addressing the drawbacks and issues, looking into potential future directions, and making use of the automation and comprehensive management capabilities of DT. The installation cost of DT technology in livestock farms should be subsidized by the government, and collaboration should be increased between the government, technology companies, engineers, farmers, and animal scientists. The government can take some initiatives to introduce and adopt DT. This can include, offering partial compensation (10-25%) or arranging low-interest credit facilities; tax reduction (up to 10%) on PLF tools and AI software; encouraging researchers by offering grants; exhibiting pilot DT model for potential stakeholders; ensuring uninterrupted internet facilities in rural areas; arranging training and workshop; fostering innovation by encouraging public-private partnerships; ensuring efficient and secure farm data sharing and, distributing awards among stakeholders for sustainable use of DT. The incorporation of DT in farm management is expected to improve animal welfare and sustainably increase livestock output with further study and conscientious application.

CONCLUSIONS AND RECOMMENDATIONS

DT assists in livestock farm operations by monitoring and predicting animal behavior, facilitating early disease detection, and controlling the farm environment remotely. However, currently, there are several challenges- high establishment costs, lack of technological knowledge, data privacy, lacking human-animal relationships, and sustainability issues. In the future, more emphasis should be given to collaborative research on developing protocols for data collection and integration, improving AI models, analyzing the cost-benefit ratio, monitoring animal behavior and welfare, evaluating environmental impact, guaranteeing cybersecurity, developing stakeholders’ training, and broadening the application area. It is expected that these kinds of efforts may improve the accuracy of the DT model, assist in investment decisions, ensure better animal behavior and welfare, improve resource management for sustainable farming, increase trust among stakeholders, raise DT adoption rates, and elevate livestock production. Governments and private sectors need to come forward with adequate funding, facilitating collaborative research, which will assist in expanding the domain of DT application in livestock farming systems.

ACKNOWLEDGMENTS

The authors would like to acknowledge the Korea Institute of Planning and Evaluation for Technology in Food, Agriculture and Forestry (IPET) and Korea Smart Farm R&D Foundation through the Smart Farm Innovation Technology Development Program, funded by the Ministry of Agriculture, Food and Rural Affairs (MAFRA) and Ministry of Science and ICT (MSIT) and Rural Development Administration (RDA) (421023-04). Also, work was supported by the Korea Institute of Planning and Evaluation for Technology in Food, Agriculture, Forestry (IPET) through Agri-Food Export Enhancement Technology Development Program, funded by Ministry of Agriculture, Food and Rural Affairs (MAFRA) (RS-2023-00231738).

NOVELTY STATEMENT

This study presents a comprehensive analysis of the literature concerning the application of digital twin technology in livestock farm management, an area that has remained relatively underexplored until now. The review opens the door for more studies and developments in the field of precision livestock management.

AUTHOR’S CONTRIBUTIONS

Conceptualization: Md Kamrul Hasan, Hong-Seok Mun, Young-Hwa Kim, Chul-Ju Yang; Methodology: Md Kamrul Hasan, Hong-Seok Mun; Writing - original draft preparation: Md Kamrul Hasan, Hong-Seok Mun, Keiven Mark Bigtasin Ampode, Young-Hwa Kim, Eddiemar Baguio Lagua, Md Sharifuzzaman; Writing - review and editing: Md Kamrul Hasan, Hong-Seok Mun, Keiven Mark Bigtasin Ampode, Young-Hwa Kim, Eddiemar Baguio Lagua, Hae-Rang Park, Md Sharifuzzaman, Chul-Ju Yang; Formal analysis: Md Kamrul Hasan, Hong-Seok Mun; Supervision: Chul-Ju Yang.

Conflicts of Interest

The authors have declared no conflict of interest.

REFERENCES

Abdolrasol MGM, Hussain SMS, Ustun TS, Sarker MR, Hannan MA, Mohamed R, Ali JA, Mekhilef S, Milad A (2021). Artificial Neural Networks Based Optimization Techniques: A Review. Electronics, 10: 2689. https://doi.org/10.3390/electronics10212689

Agnusdei GP, Elia V, Gnoni MG (2021). Is Digital Twin Technology Supporting Safety Management? A Bibliometric and Systematic Review. Appl. Sci., 11: 2767. https://doi.org/10.3390/app11062767

Ahn EY, Kim SY (2023). Digital Twin Application and Bibliometric Analysis for Digitization and Intelligence Studies in Geology and Deep Underground Research Areas. Data, 8: 73. https://doi.org/10.3390/data8040073

Aivaliotis P, Georgoulias K, Chryssolouris G (2019). The use of Digital Twin for predictive maintenance in manufacturing. Int. J. Comput. Integr. Manuf., 32: 1067–1080. https://doi.org/10.1080/0951192X.2019.1686173

Alameer A, Kyriazakis I, Dalton HA, Miller AL, Bacardit J (2020). Automatic recognition of feeding and foraging behaviour in pigs using deep learning. Biosyst. Eng., 197: 91–104. https://doi.org/10.1016/j.biosystemseng.2020.06.013

Alelyani S (2021). Detection and Evaluation of Machine Learning Bias. Appl. Sci., 11: 6271. https://doi.org/10.3390/app11146271

Alexandratos N, Bruinsma J (2012). World agriculture towards 2030/2050: the 2012 revision, ESA Working Papers 12-03. https://doi.org/10.22004/ag.econ.288998

An D, Chen Y (2021). Digital Twin Enabled Methane Emission Abatement Using Networked Mobile Sensing and Mobile Actuation. In: IEEE 1st International Conference on Digital Twins and Parallel Intelligence (DTPI), IEEE, Beijing, China, pp. 354–357. https://doi.org/10.1109/DTPI52967.2021.9540133

Arulmozhi E, Bhuje A, Moon BE, Kim HT (2021). The Application of Cameras in Precision Pig Farming: An Overview for Swine-Keeping Professionals. Animals, 11: 2343. https://doi.org/10.3390/ani11082343

Bao J, Xie Q (2022). Artificial intelligence in animal farming: A systematic literature review. J. Clean. Prod., 331: 129956. https://doi.org/10.1016/j.jclepro.2021.129956

Benjamin M, Yik S (2019). Precision Livestock Farming in Swine Welfare: A Review for Swine Practitioners. Animals, 9: 133. https://doi.org/10.3390/ani9040133

Besteiro R, Arango T, Rodríguez MR, Fernández MD, Velo R (2018). Estimation of patterns in weaned piglets’ activity using spectral analysis. Biosyst. Eng., 173: 85–92. https://doi.org/10.1016/j.biosystemseng.2017.06.014

Birhane A (2021). Algorithmic injustice: a relational ethics approach. Patterns, 2: 100205. https://doi.org/10.1016/j.patter.2021.100205

Bos JM, Bovenkerk B, Feindt PH, van Dam YK (2018). The Quantified Animal: Precision Livestock Farming and the Ethical Implications of Objectification. Food ethics, 2: 77–92. https://doi.org/10.1007/s41055-018-00029-x

Brom FWA (1997). Onherstelbaar verbeterd : biotechnologie bij dieren als een moreel probleem / Frans W.A. Brom - F.W.A. Brom Vlaamse catalogus. Available online: https://bibliotheek.be/catalogus/fwa-brom/onherstelbaar-verbeterd-biotechnologie-bij-dieren-als-een-moreel-probleem/boek/library-marc-vlacc_1821755?branch=/root/bibnet/West-Vlaanderen (accessed 20 Jun 2024)

Chakraborty J, Majumder S, Menzies T (2021). Bias in machine learning software: why? how? what to do?, In: Proceedings of the 29th ACM Joint Meeting on European Software Engineering Conference and Symposium on the Foundations of Software Engineering. Association for Computing Machinery, New York, NY, USA, pp. 429–440. https://doi.org/10.1145/3468264.3468537

da Fonseca FN, Abe JM, de Alencar Nääs I, da Silva Cordeiro AF, do Amaral FV, Ungaro HC (2020). Automatic prediction of stress in piglets (Sus scrofa) using infrared skin temperature. Comput. Electron. Agric., 168: 105148. https://doi.org/10.1016/j.compag.2019.105148

da Silveira F, Lermen FH, Amaral FG (2021). An overview of agriculture 4.0 development: Systematic review of descriptions, technologies, barriers, advantages, and disadvantages. Comput. Electron. Agric., 189: 106405. https://doi.org/10.1016/j.compag.2021.106405

Davis SE, Lasko TA, Chen G, Siew ED, Matheny ME (2017). Calibration drift in regression and machine learning models for acute kidney injury. J. Am. Med. Inform. Assoc., 24(6): 1052–1061. https://doi.org/10.1093/jamia/ocx030

Dewan A, Ay SU, Karim MN, Beyenal H (2014). Alternative power sources for remote sensors: A review. J. Power Sources, 245: 129–143. https://doi.org/10.1016/j.jpowsour.2013.06.081

Digital Twin Market (2022). By Enterprise (Large Enterprises, Medium Enterprises, Small Enterprises), By Industry (Infrastructure, Retail, Agriculture, Oil and Gas, Aerospace, Telecommunication, Automotive and Transportation, Energy and Utilities, Healthcare, Other Industries), By Application (Product Design, Product Development, Inventory Management, Predictive Maintenance, Performance Monitoring, Business Optimization, Other Applications), Global Report by Size, Industry Analysis and Forecast, 2021-2030. Available at: https://www.strategicmarketresearch.com/market-report/digital-twin-market (accessed 23 Apr 2024)

Erdélyi V, Jánosi L (2019). Digital Twin and Shadow in Smart Pork Fetteners. Int. J. Eng. Manag. Sci., 4: 515–520. https://doi.org/10.21791/IJEMS.2019.1.63

Errandonea I, Beltrán S, Arrizabalaga S (2020). Digital Twin for maintenance: A literature review. Comput. Ind., 123: 103316. https://doi.org/10.1016/j.compind.2020.103316

Farooq MS, Riaz S, Abid A, Abid K, Naeem MA (2019). A Survey on the Role of IoT in Agriculture for the Implementation of Smart Farming. IEEE Access, 7: 156237–156271. https://doi.org/10.1109/ACCESS.2019.2949703

Farooq MS, Riaz S, Helou MA, Khan FS, Abid A, Alvi A (2022). Internet of Things in Greenhouse Agriculture: A Survey on Enabling Technologies, Applications, and Protocols. IEEE Access, 10: 53374–53397. https://doi.org/10.1109/ACCESS.2022.3166634

Fuller A, Fan Z, Day C, Barlow C (2020). Digital Twin: Enabling Technologies, Challenges and Open Research. IEEE Access, 8: 108952–108971. https://doi.org/10.1109/ACCESS.2020.2998358

García R, Aguilar J, Toro M, Pinto A, Rodríguez P (2020). A systematic literature review on the use of machine learning in precision livestock farming. Comput. Electron. Agric., 179: 105826. https://doi.org/10.1016/j.compag.2020.105826

Haldar A, Mandal SN, Deb S, Roy R, Laishram M (2022). Application of Information and Electronic Technology for Best Practice Management in Livestock Production System. In: Kumar A, Kumar P, Singh SS, Trisasongko BH, Rani M (Eds.), Agriculture, Livestock Production and Aquaculture: Advances for Smallholder Farming Systems Volume 2. Springer International Publishing, Cham, pp. 173–218. https://doi.org/10.1007/978-3-030-93262-6_11

Han X, Lin Z, Clark C, Vucetic B, Lomax S (2022). AI Based Digital Twin Model for Cattle Caring. Sensors, 22: 7118. https://doi.org/10.3390/s22197118

Hristov H, Dimitrov K, Penev T (2021). Use of Infrared Thermography to Monitor the Physiological Condition of Dairy Cows. In: 12th National Conference with International Participation (ELECTRONICA), pp. 1–4. https://doi.org/10.1109/ELECTRONICA52725.2021.9513724

Jarrahi MH, Memariani A, Guha S (2023). The Principles of Data-Centric AI (DCAI). Communications of the ACM, 66: 84–92. https://doi.org/10.1145/3571724

Jeong DY, Jo SK, Lee IB, Shin H, Kim JG (2023). Digital Twin Application: Making a Virtual Pig House Toward Digital Livestock Farming. IEEE Access, 11: 121592–121602. https://doi.org/10.1109/ACCESS.2023.3313618

Jo S-K, Park D-H, Park H, Kim S-H (2018). Smart Livestock Farms Using Digital Twin: Feasibility Study. In: International Conference on Information and Communication Technology Convergence (ICTC), pp. 1461–1463. https://doi.org/10.1109/ICTC.2018.8539516

Jo S-K, Park D-H, Park H, Kwak Y, Kim S-H (2019). Energy Planning of Pigsty Using Digital Twin. In: International Conference on Information and Communication Technology Convergence (ICTC), IEEE, Jeju Island, South Korea, pp. 723–725. https://doi.org/10.1109/ICTC46691.2019.8940032

Kaklauskas A (2018). Agriculture 5.0: Smart Agriculture, AI, IoT, Society 5.0, Big Data Analytics, Blockchain, Digital Twins, Robotics, Drones and Immersive Reality. Available online: http://ir.lib.ruh.ac.lk/xmlui/handle/iruor/13735

Kashiha M, Bahr C, Ott S, Moons CPH, Niewold TA, Ödberg FO, Berckmans D (2014). Automatic weight estimation of individual pigs using image analysis. Comp. Electron. Agric., 107: 38–44. https://doi.org/10.1016/j.compag.2014.06.003

Kashiha MA, Bahr C, Ott S, Moons CPH, Niewold TA, Tuyttens F, Berckmans D (2014). Automatic monitoring of pig locomotion using image analysis. Livest. Sci., 159: 141–148. https://doi.org/10.1016/j.livsci.2013.11.007

Keates O (2019). The Design and Validation of a Process Data Analytics Methodology for Improving Meat and Livestock Value Chains. In Doctoral Consortium of the 17th International Conference on Business Process Management (BPM): Vienna. https://bpm2019.ai.wu.ac.at/

Khoramshahi E, Hietaoja J, Valros A, Yun J, Pastell M (2014). Real-time recognition of sows in video: A supervised approach. Inf. Process. Agric., 1: 73–81. https://doi.org/10.1016/j.inpa.2014.07.002

Kumar S, Choudhary S, Kumar R, Kumar A, Pal P, Mahajan K (2019). Animal Sentience and Welfare: An Overview. Int. J. Curr. Microbiol. Appl. Sci., 8: 635–646. https://doi.org/10.20546/ijcmas.2019.808.074

Lee W, Ham Y, Ban T-W, Jo O (2019). Analysis of Growth Performance in Swine Based on Machine Learning. IEEE Access, 7: 161716–161724. https://doi.org/10.1109/ACCESS.2019.2951522

Liu R, Gailhofer P, Gensch CO, Köhler A, Wolff F, Monteforte M, Urrutia C, Cihlarova P, Williams R (2019). Impacts of the digital transformation on the environment and sustainability. Issue Paper under Task 3.

Loukatos D, Arvanitis KG (2021). Multi-Modal Sensor Nodes in Experimental Scalable Agricultural IoT Application Scenarios. In: Krause, P., Xhafa, F. (Eds.), IoT-Based Intelligent Modelling for Environmental and Ecological Engineering: IoT Next Generation EcoAgro Systems. Springer International Publishing, Cham, pp. 101–128. https://doi.org/10.1007/978-3-030-71172-6_5

Lu Y, Liu C, Wang KI-K, Huang H, Xu X (2020). Digital Twin-driven smart manufacturing: Connotation, reference model, applications and research issues. Robot. Comput.-Integr. Manuf., 61: 101837. https://doi.org/10.1016/j.rcim.2019.101837

Malik P, Sneha, Garg D, Bedi H, Gehlot A, Malik PK (2023). An Improved Agriculture Farming Through the Role of Digital Twin. In: 4th International Conference on Electronics and Sustainable Communication Systems (ICESC), IEEE, Coimbatore, India, pp. 667–671. https://doi.org/10.1109/ICESC57686.2023.10193522

Mallinger K, Purcell W, Neubauer T (2022). Systemic design requirements for sustainable Digital Twins in precision livestock farming. In: Berckmans D, Oczak M, Iwersen M, Wagener K (Eds.), Precision Livestock Farming. pp. 718–725. https://doi.org/10.34726/4644

Mat I, Mohd Kassim MR, Harun AN, Yusoff IM (2018). Smart Agriculture Using Internet of Things, In: IEEE Conference on Open Systems (ICOS), IEEE, Langkawi, Malaysia, pp. 54–59. https://doi.org/10.1109/ICOS.2018.8632817

Mhapsekar RU, O’Shea N, Davy S, Abraham L (2024). Hybrid Blended Deep Learning Approach for Milk Quality Analysis. IEEE Trans. Emerg. Top. Comput. Intell., 8: 2253–2268. https://doi.org/10.1109/TETCI.2024.3369331

Mu M, Zhou Y, Wu D (2022). Digital Twins on Animal Husbandry: Insights and Application. Procedia Computer Science, 9th International Conference on Information Technology and Quantitative Management, 214: 1182–1189. https://doi.org/10.1016/j.procs.2022.11.294

Murali D, Acosta-Pazmiño IP, Loris A, García AC, Benni S, Tinti F, Gomes J (2024). Experimental assessment of a solar photovoltaic-thermal system in a livestock farm in Italy. Sol. Energy Adv., 100051. https://doi.org/10.1016/j.seja.2024.100051

Nasirahmadi A, Hensel O (2022). Toward the Next Generation of Digitalization in Agriculture Based on Digital Twin Paradigm. Sensors, 22(2): 498. https://doi.org/10.3390/s22020498

Nasirahmadi A, Sturm B, Edwards S, Jeppsson K-H, Olsson A-C, Müller S, Hensel O (2019). Deep Learning and Machine Vision Approaches for Posture Detection of Individual Pigs. Sensors, 19: 3738. https://doi.org/10.3390/s19173738

Neethirajan S (2023). The Significance and Ethics of Digital Livestock Farming. AgriEngineering, 5: 488–505. https://doi.org/10.3390/agriengineering5010032

Neethirajan S (2022). Affective State Recognition in Livestock—Artificial Intelligence Approaches. Animals, 12: 759. https://doi.org/10.3390/ani12060759

Neethirajan S (2020). Transforming the Adaptation Physiology of Farm Animals through Sensors. Animals, 10: 1512. https://doi.org/10.3390/ani10091512

Neethirajan S, Kemp B (2021). Digital Twins in Livestock Farming. Animals, 11: 1008. https://doi.org/10.3390/ani11041008

Oudenhoven JFM, Vullers RJM, van Schaijk R (2012). A review of the present situation and future developments of micro-batteries for wireless autonomous sensor systems. Int. J. Energy Res., 36: 1139–1150. https://doi.org/10.1002/er.2949

Page MJ, McKenzie JE, Bossuyt PM, Boutron I, Hoffmann TC, Mulrow CD, Shamseer L, Tetzlaff JM, Akl EA, Brennan SE, Chou R, Glanville J, Grimshaw JM, Hróbjartsson A, Lalu MM, Li T, Loder EW, Mayo-Wilson E, McDonald S, McGuinness LA, Stewart LA, Thomas J, Tricco AC, Welch VA, Whiting P, Moher D (2021). The PRISMA 2020 statement: an updated guideline for reporting systematic reviews. BMJ, 372: n71. https://doi.org/10.1136/bmj.n71

Pandey S, Kalwa U, Kong T, Guo B, Gauger PC, Peters DJ, Yoon K-J (2021). Behavioral Monitoring Tool for Pig Farmers: Ear Tag Sensors, Machine Intelligence, and Technology Adoption Roadmap. Animals, 11: 2665. https://doi.org/10.3390/ani11092665

Paraforos DS, Griepentrog HW (2021). Digital Farming and Field Robotics: Internet of Things, Cloud Computing, and Big Data. In: Karkee M, Zhang Q (Eds.), Fundamentals of Agricultural and Field Robotics. Springer International Publishing, Cham, pp. 365–385. https://doi.org/10.1007/978-3-030-70400-1_14

Peladarinos N, Piromalis D, Cheimaras V, Tserepas E, Munteanu RA, Papageorgas P (2023). Enhancing Smart Agriculture by Implementing Digital Twins: A Comprehensive Review. Sensors, 23: 7128. https://doi.org/10.3390/s23167128

Petrov P, Atanasova T (2021). Digital Twins with Application of AR and VR in Livestock Instructions. Probl. Eng. Cybern. Robot., 77. https://doi.org/10.7546/PECR.77.21.05

Pezzuolo A, Milani V, Zhu D, Guo H, Guercini S, Marinello F (2018). On-Barn Pig Weight Estimation Based on Body Measurements by Structure-from-Motion (SfM). Sensors, 18: 3603. https://doi.org/10.3390/s18113603

Platform Industrie 4.0 (2024). Available at: https://www.plattform-i40.de/IP/Navigation/EN/Home/home.html (accessed 16 Jun 2024).

Prior F, Almeida J, Kathiravelu P, Kurc T, Smith K, Fitzgerald TJ, Saltz J (2020). Open access image repositories: high-quality data to enable machine learning research. Clin. Radiol., 75: 7–12. https://doi.org/10.1016/j.crad.2019.04.002

Proctor H (2012). Animal Sentience: Where Are We and Where Are We Heading? Animals, 2: 628–639. https://doi.org/10.3390/ani2040628

Proctor HS, Carder G, Cornish AR (2013). Searching for Animal Sentience: A Systematic Review of the Scientific Literature. Animals, 3: 882–906. https://doi.org/10.3390/ani3030882

Pylianidis C, Osinga S, Athanasiadis IN (2021). Introducing digital twins to agriculture. Comput. Electron. Agric., 184: 105942. https://doi.org/10.1016/j.compag.2020.105942

Raba D, Tordecilla RD, Copado P, Juan AA, Mount D (2022). A Digital Twin for Decision Making on Livestock Feeding. Inf. J. Appl. Anal., 52: 267–282. https://doi.org/10.1287/inte.2021.1110

Roubík H, Mazancová J (2020). Suitability of small-scale biogas systems based on livestock manure for the rural areas of Sumatra. Environ. Dev., 33: 100505. https://doi.org/10.1016/j.envdev.2020.100505

Saha CK, Nandi R, Rahman MA, Alam MM, Møller HB (2024). Biogas technology in commercial poultry and dairy farms of Bangladesh: present scenario and future prospect. Biomass Convers. Biorefinery, 2024; 14(7): 8407-8418. https://doi.org/10.1007/s13399-022-02938-8

Saidu A, Clarkson AM, Adamu SH, Mohammed M, Jibo I (2017). Application of ICT in Agriculture: Opportunities and Challenges in Developing Countries. Int. J. Comput. Sci. Math. Theory, 3(1): 8-18.

Salazar-Moya A, Garcia MV (2021). Lot Streaming in Different Types of Production Processes: A PRISMA Systematic Review. Designs, 5(4): 67. https://doi.org/10.3390/designs5040067

Seibel H, Weirup L, Schulz C (2020). Fish Welfare – Between Regulations, Scientific Facts and Human Perception. Food ethics, 5(1): 4. https://doi.org/10.1007/s41055-019-00063-3

Smith MJ (2018). Getting value from artificial intelligence in agriculture. Anim. Prod. Sci., 60: 46–54. https://doi.org/10.1071/AN18522

Sreedevi TR, Santosh Kumar MB (2020). Digital Twin in Smart Farming: A Categorical Literature Review and Exploring Possibilities in Hydroponics. In: Advanced Computing and Communication Technologies for High Performance Applications (ACCTHPA), Cochin, India, pp. 120–124. https://doi.org/10.1109/ACCTHPA49271.2020.9213235

Tarek N, El-Hefnawy NA, Abdel-Kader H, Abdelatey A (2022). A Flexible Digital Twin Framework for Poultry Feed Industrial Factory. In: 32nd International Conference on Computer Theory and Applications (ICCTA), IEEE, Alexandria, Egypt, pp. 253–257. https://doi.org/10.1109/ICCTA58027.2022.10206352

Tilman D, Balzer C, Hill J, Befort BL (2011). Global food demand and the sustainable intensification of agriculture. Proc. Natl. Acad. Sci., 108: 20260–20264. https://doi.org/10.1073/pnas.1116437108

Top 10 Strategic Technology Trends for 2019: Digital Twins (2024). Available at: https://www.gartner.com/en/documents/3904569 (accessed 23 Apr 2024).

Tzachor A, Richards CE, Jeen S (2022). Transforming agrifood production systems and supply chains with digital twins. Npj Sci. Food, 6(1): 47. https://doi.org/10.1038/s41538-022-00162-2

Upadhyay AK, Khandelwal K (2019). Artificial intelligence-based training learning from application. Dev. Learn. Organ. Int. J., 33: 20–23. https://doi.org/10.1108/DLO-05-2018-0058

Valero MR, Hicks BJ, Nassehi A (2023). A Conceptual Framework of a Digital-Twin for a Circular Meat Supply Chain. In: Kim K-Y, Monplaisir L, Rickli J (Eds.), Flexible Automation and Intelligent Manufacturing: The Human-Data-Technology Nexus, Lecture Notes in Mechanical Engineering. Springer International Publishing, Cham, pp. 188–196. https://doi.org/10.1007/978-3-031-18326-3_19

Verdouw C, Tekinerdogan B, Beulens A, Wolfert S (2021). Digital twins in smart farming. Agric. Syst., 189: 103046. https://doi.org/10.1016/j.agsy.2020.103046

White RP, Schofield CP, Green DM, Parsons DJ, Whittemore CT (2004). The effectiveness of a visual image analysis (VIA) system for monitoring the performance of growing/finishing pigs. Anim. Sci., 78: 409–418. https://doi.org/10.1017/S1357729800058811

World Population Prospects 2022: Summary of Results (2022). Available at: https://www.un.org/development/desa/pd/content/World-Population-Prospects-2022 (accessed 23 Apr 2024)

Zhang C, Campana P, Yang J, Zhang J, Yan J (2017). Can Solar Energy be an Alternative Choice of Milk Production in Dairy Farms? A Case study of Integrated PVWP System with Alfalfa and Milk Production in Dairy Farms in China. In: Energy Procedia, 8th International Conference on Applied Energy, ICAE2016, Beijing, China, pp. 3953–3959. https://doi.org/10.1016/j.egypro.2017.03.822

Zhang Q, Yang LT, Chen Z, Li P (2018). A survey on deep learning for big data. Inf. Fusion, 42: 146–157. https://doi.org/10.1016/j.inffus.2017.10.006

Zhang W, Gu X, Tang L, Yin Y, Liu D, Zhang Y (2022). Application of machine learning, deep learning and optimization algorithms in geoengineering and geoscience: Comprehensive review and future challenge. Gondwana Res., 109: 1–17. https://doi.org/10.1016/j.gr.2022.03.015

Zhang X, Cao Z, Dong W (2022). Overview of Edge Computing in the Agricultural Internet of Things: Key Technologies, Applications, Challenges. IEEE Access, 8: 141748–141761. https://doi.org/10.1109/ACCESS.2020.3013005

Zhang Yi, Zhang Yu, Gao M, Dai B, Kou S, Wang X, Fu X, Shen W (2023). Digital Twin Perception and Modeling Method for Feeding Behavior of Dairy Cows. Comput. Electron. Agric., 214: 108181. https://doi.org/10.2139/ssrn.4463418