The Impact of 6G-IoT Technologies on the Development of Agriculture 5.0: A Review

Round 1

Reviewer 1 Report

Comments for author File: Comments.pdf

Author Response

- Reviewer #1

This manuscript presents interesting reviews on the lately developed information technology and its impacts on agriculture. The content is insightful. However, this manuscript is more like an technical outlook rather than an academic reviews. The 5G is only getting start in daily lift and few meaningful/influential applications has been introduced to agricultural field. 6G is still a concept. So it’s kind of strange to see this manuscript. Nevertheless, this manuscript is well written and insightful.

               We would like to thank the reviewer for his positive assessment of our work.

To improve it to the level of academical paper, I have several questions/advice for authors.

1. Should this manuscript be classified as article? It’s kind of confusing. Obviously, it’s not a research article and appears like a review.

We would like to thank the reviewer for his constructive comment. Indeed, our paper is a review, as we aim to combine information from multiple sources in order to provide an overview of the anticipated impact of 6G-IoT technologies on the future development of agriculture. Therefore, we changed the title to

"The Impact of 6G-IoT Technologies on the Development of Agriculture 5.0: A Review."

2. In the abstract section, agriculture 2.0 is missing.

As suggested by the reviewer, we added a reference for Agriculture 2.0 to the abstract section in lines 2-4 as follows:

“Over the years, farming methods have evolved significantly, progressing from Agriculture 1.0, which relied on primitive tools, to Agriculture 2.0, which incorporated machinery and advanced farming practices, and subsequently to Agriculture 3.0, which emphasized mechanization and employed intelligent machinery and technology to enhance productivity levels.”.

3. L48-57, Even though 5G is well established framework, applications based on 5G is scarce. Authors want to highlight the 6G, but both are only means not ends. The value of information technology in agriculture hasn’t been justified as they didn’t increase the yields/profits as greatly as the other revolution like genetic breading, large farm machinery etc.

We agree that both 5G and 6G networks are means to achieving the aims of modern agriculture, and the reviewer is correct to point out that revolutions are also occuring in other scientific areas like genetic breeding. However, even in these cases, ICT remains significant because it serves as the vehicle for speeding up research by easing the burdens of data collection and management, genetic analysis, and predictive modeling to assess the effectiveness of various breeding strategies, thereby allowing breeders to make better decisions, and speeding up the rate at which plants and animals are improved genetically.

During the initial phases of its development, smart farming has indeed encountered problems, and in certain cases, the deployed systems failed to produce the anticipated outcomes. However, significant progress has been made in recent years, and ICT has demonstrated its value in the agricultural sector. The most reliable assessment of ICT's impact on agriculture can be found in the Global Smart Farming Market Forecast, which predicts the market will grow from USD 12.80 billion in 2021 to USD 33.69 billion in 2029 (ref. to [26]).

 The text at the end of Section 2.1 was revised and expanded to emphasize this point:

“Farmers can employ data-driven agriculture techniques for a variety of purposes, including strategic planning, real-time monitoring, performance evaluation, predictive forecasting, optimization, and effective event management. However, the widespread adoption of these techniques has been hampered by a number of obstacles, such as limited connectivity solutions and expensive and, in some cases, inaccurate data collection. Despite early obstacles, smart farming has made significant advances in recent years, with Information and Communication Technology (ICT) demonstrating its value in the agricultural sector [25]. Overall, adopting data-driven strategies has proven to increase agricultural productivity by increasing crop yields, reducing input costs, and minimizing crop loss. According to the Global Smart Farming Market Forecast, the market is anticipated to experience substantial growth, with projections estimating its value to rise from USD 12.80 billion in 2021 to USD 33.69 billion by 2029 [26], providing a reliable assessment of the expected impact of ICT on agriculture.”

4. How the definitions of agriculture 1.0, 2.0, … are made? I’m not sure of its correctness but I can’t see the significance of the agriculture 3.0 and 4.0. There are little revolutions in them. Lot of them are just fancy ideas rather than practical changes. For example, even though Variable Rate Technology was proposed long time ago that combines sensor with variable control methods to save agrochemicals, it’s not appealing to farmers as the revenues can’t cover inputs. In contrast, it’s more appealing to farmers to have reliable machinery. From my point of view, it’s better to divide the time according to paradigm shift. Perhaps this is not done by authors. Discussion is advised.

The division of time according to paradigm shifts can indeed provide a valuable framework for discussion and analysis. However, since our work revolves around technological breakthroughs, we find it more suitable to follow the evolutionary path of ICT technology, as certain ideas, though initially seeming like fanciful concepts, often serve as steppingstones for future development and refinement.

To clarify this point of view, we added the following statement at the beginning of Section 2.1:

"Following a technological evolution-based classification of agriculture eras as proposed in the related scientific literature [17], [18], we can identify four distinct eras: Agriculture 1.0, Agriculture 2.0, Agriculture 3.0, and Agriculture 4.0.".

5. L253-265, to include the latest development in this field, below are for your reference
   • Yu J, Zhang J, Shu A, et al. Study of convolutional neural network-based semantic segmentation methods on edge intelligence devices for field agricultural robot navigation line extraction[J]. Computers and Electronics in Agriculture, 2023, 209: 107811.
   • Zhang Y, Yu J, Chen Y, et al. Real-time strawberry detection using deep neural networks on embedded system (rtsd-net): An edge AI application[J]. Computers and Electronics in Agriculture, 2022, 192: 106586.

These are undoubtedly interesting applications, and we included them as follows in subsection 3.1 ("Edge Computing and Pervasive Artificial Intelligence"):

“Edge AI applications have emerged as a significant area of research, demonstrating their effectiveness in a variety of agricultural applications. For instance, cutting-edge computer vision methods utilizing deep learning can facilitate the extraction of navigation lines in intricate farmland environments for field robots [57]. Moreover, the integration of lightweight neural networks and edge computing enables real-time crop detection by harvesting robots [58].”  

6. Edge AI will reduce the need for high throughput wireless communications. What the authors think about this issue?

One of the primary advantages of Edge AI is its ability to process data locally, rather than sending every piece of data to a central server or the cloud for processing, which lowers bandwidth requirements. Edge AI will therefore lessen the requirement for high-throughput wireless communications.

However, Edge AI is not a solution that fits all needs, and there will be situations in which communication with a central server or cloud is required. Complex processing tasks that require substantial computational resources often necessitate offloading the workload to the cloud. Additionally, situations that call for collaborative data sharing, advanced analytics, or the need for centralized data insights and decision-making may also require communication with the cloud. Therefore, the extent to which edge AI reduces the need for high-throughput wireless communications depends on several factors, including the specific application requirements, the nature of the data being processed, and the capabilities of the edge devices involved.

To clarify this point, the following has been added to subsection 3.1 ("Edge Computing and Pervasive Artificial Intelligence"):

“Overall, edge computing and edge AI offer the benefits of local data processing, which reduces the need to send data to the cloud. This, in turn, helps to alleviate the traffic load on the backbone network. However, it is important to note that edge computing and edge AI cannot be universally applied to all deployment scenarios. They are primarily designed to provide immediate responses for real-time applications, such as navigation, coordination, and collaboration of autonomous vehicles and robots in farming operations. On the other hand, they are not well-suited for data-intensive tasks like large-scale analytics, high-resolution image processing, and augmented reality applications. Consequently, for complex tasks that require significant computational resources, it is still necessary to transmit data to the cloud. The extent to which edge computing and edge AI can reduce the need for communication with the cloud is dependent on the requirements of the specific application, the nature of the data, and the capabilities of the edge device.”

7. L259, AI models are not trained on edge devices but deployed on edge devices.

Thank you for allowing us to clarify this point.

Your remark is generally accurate and reflects the standard practice in the field of artificial intelligence, where models are trained on powerful hardware and then deployed to edge devices. However, in our study, we present the future of Edge AI in which Federated Learning is utilized and, as a result, the training process is decentralized, and the training data is stored on the edge devices where it was generated. The edge devices participate in the training process by sending their local model updates (computed using their local data) to the central server. The central server collects updates from multiple devices, modifies the global model, and sends the updated model back to the edge devices. This process is repeated until the global model reaches the desired performance level.

We acknowledge that the respective part of section 3.1 may not effectively convey the above explanation. In response, we have rephrased it as following:

“The future of Edge AI also involves a shift in the traditional approach of training models on powerful servers and then deploying them to edge devices. With the advent of Federated Learning (FL), the training process becomes decentralized and the training data is stored on the edge devices where it originated [16]. In this approach, edge devices contribute to the training process by sending local model updates computed using their own data to a central server. The central server aggregates updates from multiple devices, modifies the global model, and then distributes the updated version to the edge devices. This iterative procedure continues until the global model reaches the desired performance level. As a result, the need for data transmission from the network’s edge to the cloud is significantly reduced, leading to increased security and an accelerated training process [59, 60].”

8. With all due respect, to what extent is ChatGPT used in writing? Comparing to research article, it’s hard to determine the academic contribution of a review manuscript.

In academic writing, a research article and a review article serve distinct purposes. A research article is intended to present the results of original research, whereas a review article provides a comprehensive overview and analysis of existing research on a particular topic. A review article plays a crucial role in assisting researchers and readers in gaining a comprehensive understanding of the current state of knowledge, identifying research gaps, and identifying possible paths for future research. Therefore, review articles are essential for staying current and knowledgeable in a particular field.

We believe that our work falls within this category as it presents and discusses various ongoing and anticipated technological developments as well as their impact on future agriculture.

Regarding ChatGPT, it indeed has impressive capabilities; however, at this point, it cannot be used to create a review article, which is obvious from the following statement on the respective site:

“ChatGPT is not connected to the internet, and it can occasionally produce incorrect answers. It has limited knowledge of world and events after 2021 and may also occasionally produce harmful instructions or biased content.”

Maybe someday ChatGPT will be able to produce scientific work, but for now it doesn't have the required data and, more importantly, the critical analysis capabilities. 

9. L432, even though 5G is famous for its reliable latency communication, the low power wide area network, known as NB-IoT is part of it protocol stack. And it’s more interesting and profitable.

Indeed, NB-IoT as well as LTE-M are the way that 3GPP responded to the need for cellular low power wide area networks for IoT devices. They are both developed to utilize the infrastructure of 4G LTE-A networks and support simple narrowband IoT devices with low capabilities. 3GPP eventually incorporated both into the 5G standard, ensuring continued support for these technologies. Throughout the entire paper, we refer to 5G technology as a whole concept, including NB-IoT.

While NB-IoT, LTE-M are, more or less, adjusted to the current needs of smart agriculture they still are scaled down versions of 4G LTE and cannot possibly cover the requirements of future smart agriculture as they are discussed at subsection 2.2.

10. Even though the section 4 listed several fancy possible applications of 6G, I seriously doubt the feasibility and practicality of them. For example, how the AR improve field disease assessment and reduce agrochemicals?

As we previously mentioned, one of the objectives of a review paper is to identify all possible paths for future research. Therefore, indeed, some of them may prove to be impractical, at least for specific use cases, but this cannot be determined until the respective research path is explored.

Therefore, in our work we simply identify that there is a research community that considers AR as a valuable tool and there are corresponding published works to support the viability of this research path.

To further enhance the specific section, we added references to two more publications:

• Garg, Simran, Priya Sinha, and Archana Singh. "Overview of augmented reality and its trends in agriculture industry." In IOT with Smart Systems: Proceedings of ICTIS 2021, Volume 2, pp. 627-636. Springer Singapore, 2022.
• Mahenthiran, Nitharshana, Haarini Sittampalam, Sinthumai Yogarajah, Sathurshana Jeyarajah, Sanjeevi Chandrasiri, and Archchana Kugathasan. "Smart Pest Management: An Augmented Reality-Based Approach for an Organic Cultivation." In 2021 2nd International Informatics and Software Engineering Conference (IISEC), pp. 1-6. IEEE, 2021.

The respective text in subsection 4.5 ("Enhanced Remote Disease Assessment and Augmented Reality") is extensively rewritten as follows:

“Furthermore, Augmented Reality (AR) can aid farmers in numerous ways. AR is a technology that superimposes computer-generated images onto the real world, enhancing the objects and information in the user’s view. Thus, AR can significantly improve field disease assessment and reduce the use of agrochemicals by facilitating accurate pest identification and detection of pest attacks and infections in early stages, as it can provide an immersive and visual representation of the pest’s life cycle stages [124]. Thus, AR will facilitate the application of the appropriate treatment at the appropriate time, reducing the need for excessive agrochemical use.

Moreover, by combining digital information with a farmer’s environment in real time, augmented reality is able to deliver useful information such as the condition of crops and machinery, weather updates, soil and water conditions, and AI-based plant disease diagnosis via smartphones or glasses, enabling farmers to make informed decisions [125,126]. In addition, AR agricultural applications will benefit from the secure, reliable, and low-latency connectivity of 6G networks. This will enable them to transfer data in real time without any lag that could cause motion sickness to the end-user and reduce the efficiency of the AR application. Finally, in education and knowledge sharing, AR can be used as an educational tool to train farmers and agricultural workers about pest identification, prevention methods, and organic farming practices [127,128].”.

In summary, this manuscript is interesting. It may attract readers outside of agricultural engineers.

We would like to thank the reviewer for his positive assessment of our work.

Author Response File: Author Response.pdf

Reviewer 2 Report

The paper clearly provides contribution by a review of the current state of research and explaining some of the challenges in regard of the impact of 6G-IoT Technologies on the Development of so called Agriculture 5.0.

The paper introduces the scope of Agriculture 5.0 and technologies that will be leveraged in the 6G IoT communication systems. It highlights the importance and influence of these developing technologies in the further advancement of smart agriculture and conclude with a discussion of future challenges and opportunities.

This paper is well organized and provides extensive information about the current research literature overview. No specific research weaknesses are found.

Most of the references have a year of publication typed in bold font, but some of them (for example 32, 34, 39, 53, 73, etc.) are not. Authors should make the text uniform. Also, minor revisions of the English language are required.

Author Response

-Reviewer #2

The paper clearly provides contribution by a review of the current state of research and explaining some of the challenges in regard of the impact of 6G-IoT Technologies on the Development of so called Agriculture 5.0. The paper briefly introduces the scope of Agriculture 5.0 and technologies that will be leveraged in the 6G IoT communication systems. It highlights the importance and influence of these developing technologies in the further advancement of smart agriculture and conclude with a discussion of future challenges and opportunities. This paper is well organized and provides extensive information about the current research literature overview.

               We would like to thank the reviewer for his positive assessment and thorough analysis of our work.

The work should be classified as a review paper, as it does not provide original research results. 

We would like to thank the reviewer for his constructive comment. Indeed, our paper is a review, as we aim to combine information from multiple sources in order to provide an overview of the anticipated impact of 6G-IoT technologies on the future development of agriculture. Therefore, we changed the title to

"The Impact of 6G-IoT Technologies on the Development of Agriculture 5.0: A Review."

In the paper a number of modern technologies is mentioned as applicable to the topic, but their introduction and application are not sufficiently explained. For example, from the text in 3.4. it is unclear why quantum sensing (QS), communication and computation are introduced and related to the topic of the smart agriculture. Quantum sensing has rather specific applications, and the authors are encouraged to either add an extensive explanation and references of QS application that could be used in agriculture, or dismiss this idea at all. The sentence "detect even the slightest changes in an agricultural system" should be significantly expanded.  

We acknowledge that certain parts of the paper required enhancement. Consequently, we have rewritten and expanded several sections (i.e., Abstract, Introduction, 2.1., 3.1, 3.3., 3.4., 3.5., 4.1., 4.4., 4.5., 4.6.)

Furthermore, we believe that one of the objectives of a review paper is to identify all possible future research directions, and as such we refer to quantum sensing (QS), communication, and computation. Nevertheless, we agree with the reviewer that the specific topic and the manner that it was described in the original text required improvement. Consequently, we revised all relevant sections as follows:

Regarding Quantum Sensing we first improved the description of the quantum sensor at subsection 3.4.:

“Many elementary quantum systems, including atomic spin systems, NV (nitrogen-vacancy) center ensembles, and trapped ions, can serve as quantum sensors [69]. Due to the extreme sensitivity of quantum states to even the smallest changes, these sensors have a higher sensitivity than conventional sensors, and they have the potential for more precise monitoring of physical parameters such as electromagnetic fields and temperature.”

We also amended Section 4.1 to include QS applications and revised the phrase "detect even the most minute changes in an agricultural system." Thus, the respective text is updated as follows:

“Additionally, the utilization of quantum sensing technology can contribute significantly to this objective. Quantum sensors, relying on highly sensitive quantum systems, offer greater sensitivity compared to conventional sensors. Thus, two recent research studies [104], [105] propose the use of quantum sensors for the precise measurement of Photosynthetically Active Radiation (PAR), which is regarded as one of the most important environmental metrics required to evaluate plant photosynthesis. Furthermore, in [106], the authors employed devices equipped with multiple quantum sensors in order to precisely measure PAR as part of a cost-effective UAV-based thermal system to generate predicted maps of Leaf Water Potential (LWP). Moreover, quantum sensors can be employed to accurately measure pH levels and ions, enabling the precise assessment of essential nutrients such as nitrogen, phosphorus, and potassium, which facilitates the implementation of precise fertilization strategies. Additionally, the deployment of multiple quantum sensors in a network allows for the utilization of inter-sensor correlations, further enhancing the overall sensing capabilities of the system [107]. Moreover, [108] investigates the use of quantum dots, which are typically nanoscale semiconductor particles or structures with unique quantum mechanical properties due to their size, for the production of nanobiosensors that are able to detect distinct targets in plants, such as pathogens, nutrients, and pesticides.”

Regarding quantum computing, we improved the description of its capabilities at subsection 3.4:

“The main advantage of quantum computing is quantum parallelism, which enables quantum computers to simultaneously compute multiple outputs of a function, exceeding the capabilities of classical computing. Quantum computing is anticipated to be widely utilized for accelerating the analysis of big data sets and minimizing the training phase of machine learning algorithms. By going beyond classical binary computing, quantum computing can significantly reduce the training and execution times of machine learning algorithms, making them suitable for real-time, computationally demanding applications. The new field of research that emerges from the combination of quantum computing with machine learning is referred to as Quantum Machine Learning (QML) [72], [73]”

               We also enhanced subsection 4.4 by adding the following:

               “Furthermore, by harnessing the parallel processing capability of quantum computing, which is expected to deliver significantly increased processing speeds compared to traditional computing once fully developed, the time required for training and executing AI algorithms can be greatly reduced [72,73,116]. Several scientific studies explore the use of quantum computing [117], [118] and quantum-assisted machine learning [119], [120] for various smart agriculture-related objectives. These objectives include optimizing data analysis and decision-making , specifically in areas such as production planning and insurance risk assessment, enhancing image processing techniques for disease diagnosis as well as enhancing disease classification techniques. In addition, quantum computing enables the execution of complex simulations, which can be used, for example, in forecasting and modeling weather patterns. Moreover, quantum-assisted machine learning techniques can be used to predict phenotypes from genomic data and facilitate drug discovery by identifying, validating, and characterizing biological targets that are effective against disease mechanisms.”

We also revised subsection "3.4. Quantum Sensing, Communication, and Computation" to better explain the advantages of quantum communication:

“Moreover, quantum communication is expected to revolutionize data transmission by utilizing the principles of quantum mechanics to ensure a secure and tamper-proof exchange of information. Quantum communication protocols have the ability to detect any interception or eavesdropping attempts, as the act of measuring quantum states disrupts their delicate properties. Furthermore, the implementation of Quantum Key Distribution (QKD), a method that leverages quantum mechanics for securely creating and exchanging encryption keys between two parties, will further enhance the protection of sensitive agricultural data. Consequently, valuable data, including crop yields, market predictions, and sensor-derived information from the field, will be safeguarded, ensuring their integrity, confidentiality, and authenticity while mitigating the risks associated with unauthorized access and manipulation [70],[71].”

               We additionally revised subsection “4.6. Enabling Secure and Transparent Transactions” as follows:

               “In the Agriculture 5.0 ecosystem, a decentralized system for secure transactions can be developed by combining the strengths of quantum communication, AI, digital twins, and blockchain technology with the 6G-IoT networking infrastructure [121]. In such a system, quantum communication, utilizing quantum mechanics principles, can provide tamper-proof communication channels, while the integration of quantum communication and blockchain facilitates efficient, accountable, and transparent transactions. Moreover, the additional employment of AI and digital twins enables timely identification of threats, enabling proactive measures to be taken.”

               And in section “4.4. More Accurate and Reliable AI-based Applications”:

               “The combination of blockchain and quantum communication will eliminate the possibility of data manipulation by ensuring that critical data such as meteorological conditions, agricultural yields, and soil quality are securely delivered, maintained, and verified on the blockchain network. Consequently, the combination of AI, quantum communication, and blockchain will ensure that AI models are trained with data that is both reliable and immune to malicious tampering [62, 121].”

The paper insists on high bandwidth and low latency communications (for example THz communication solutions), but it does not explain why is it important, especially when taking info consideration edge computing power and on-site AI applications.

Indeed, edge computing and edge AI are transformative technologies that enable the processing and decision-making of data at or near the edge of the network. By leveraging decentralized processing and analysis, these technologies reduce reliance on cloud-based resources. They offer immediate responses when required, benefiting real-time applications such as navigation, coordination, and collaboration of autonomous vehicles and robots in farming operations.

Although edge computing resources may be deployed closer to the field than a central cloud server, fields are often located in remote and sometimes hard-to-access areas. In such cases, edge computing devices may be placed at a locally central location that satisfies key requirements such as a reliable power supply, adequate air conditioning, and robust physical security systems. Therefore, a high-bandwidth, low-latency connection is required in this scenario between the remote field and the edge computing equipment.

Furthermore, edge computing is not designed for data-intensive applications like large-scale analytics. Therefore, a large portion of the collected field data should be timely transmitted to a central cloud that has adequate computational and storage capabilities. This allows the farmer to access real-time data visualizations, receive alert reporting, and perform remote management of all his fields.

Additionally, high bandwidth and low latency communications are essential for rapid data transmission and responsive system feedback, especially for high-resolution imaging and augmented reality applications. These applications enable remote crop growth evaluation, yield identification, and early-stage disease and insect infestation detection. Delays in system responses can be especially detrimental to the effectiveness of augmented reality applications, where any perceptible delay between user actions and system responses can negatively impact the efficiency of the application.

Furthermore, as smart agriculture evolves, the number of IoT devices and the volume of generated data will continue to increase. High-bandwidth and low-latency communications provide a foundation for scalability, accommodating the growing demands of data transmission and processing in the agricultural landscape. Planning for these requirements in advance makes it easier to scale smart agriculture deployments without encountering bottlenecks or performance limitations.

In order to address this comment, we largely rewrote subsection "3.1. Edge Computing and Pervasive Artificial Intelligence", where after a discussion on edge computing and edge AI, we conclude with the following:

“Overall, edge computing and edge AI offer the benefits of local data processing, which reduces the need to send data to the cloud. This, in turn, helps to alleviate the traffic load on the backbone network. However, it is important to note that edge computing and edge AI cannot be universally applied to all deployment scenarios. They are primarily designed to provide immediate responses for real-time applications, such as navigation, coordination, and collaboration of autonomous vehicles and robots in farming operations. On the other hand, they are not well-suited for data-intensive tasks like large-scale analytics, high-resolution image processing, and augmented reality applications. Consequently, for complex tasks that require significant computational resources, it is still necessary to transmit data to the cloud. The extent to which edge computing and edge AI can reduce the need for communication with the cloud is dependent on the requirements of the specific application, the nature of the data, and the capabilities of the edge device.”

Furthermore, in subsection "4.5. Enhanced Remote Disease Assessment and Augmented Reality", which was also largely revised, we revised/added the following:

“Furthermore, the 6G networking environment will also offer higher data rates and a higher quality of data, which can include, for example, low-latency hyperspectral video streaming the detection of specific materials, identification of chemical compositions, and analysis of fine-grained details that are not visible to the human eye, while at the same time ensuring the security of the communication links.”

“In addition, AR agricultural applications will benefit from the secure, reliable, and low-latency connectivity of 6G networks. This will enable them to transfer data in real time without any lag that could cause motion sickness to the end-user and reduce the efficiency of the AR application.”

Finally, in subsection "3.5. Terra-Hertz Communications", we added the following:

“As the number of IoT devices and the volume of generated data continue to increase, the development of high-bandwidth and low-latency communication links will provide a foundation for scalability, accommodating the growing demands of data transmission and processing.”

The existing communication technologies such are Sigfox, LoRa, NB-IoT are still not used widely and effectively, so why rush into new technologies? These questions should be considered and discussion added into the work.  

We agree with the reviewer that Sigfox, LoRa, NB-IoT, as well as LTE-M are low-power narrowband communication networks that are more or less tailored to the needs of the current smart agricultural deployments, where the amount of gathered field data is relatively low.

The objective of this paper however is not to propose rushing to any new technology but to identify all possible future research paths in order to give researchers an overview and analysis of possible research directions.

But research and subsequent standardization take time. For example, the standardization of LTE (Long-Term Evolution), which serves as the foundation for NB-IoT, began in 2004 and was completed in 2008, while the standardization of NB-IoT was finalized in June 2016 and is still not widely deployed.

In our study, we discuss technologies whose initial phase of standardization is anticipated to conclude around 2030. Nevertheless, the research community will need to devote a significant amount of time and effort to achieve this objective. Therefore, from this perspective, the 2030s are considered to be close.

In order to address this comment, we added the following to the introductory section:

“Although Sigfox, LoRa, NB-IoT, and LTE-M are low-power narrowband communication networks that are, more or less, tailored to the needs of current smart agricultural deployments where the amount of gathered field data is relatively low, they will face challenges in meeting the extensive requirements of future IoT applications in the coming decade. Therefore, in this study, we aim to provide researchers with a comprehensive overview of existing research and analyze possible research directions.”

On the other hand, Internet Of Underground Things (IoUTs) and wireless underground sensor networks (WUSNs), consisting of sensors and communication devices, partly or completely buried underground for real-time soil sensing and monitoring, emerge from the needs of precision agriculture.  This topic has to be addressed if related to term Agriculture 5.0.  

Indeed, the Internet of Underground Things (IoUTs) and Wireless Underground Sensor Networks (WUSNs) are new technologies relevant to Agriculture 5.0 to the extent that, through the use of sensors and devices embedded in the soil, farmers can collect data on various parameters like moisture levels, nutrient content, and pH levels. As long as these devices utilize overground antennas for their communication, they could use the same 6G infrastructure as any other wireless IoT device.

The same holds true for underground communication, with the additional difficulty of increased signal attenuation due to soil layer interference on the transmission path. Consequently, the primary research challenge is addressing the variable path loss caused by varying soil conditions. This in turn impacts network topology, as there is a tradeoff between the energy consumption and communication range of buried sensor nodes.

To address this comment, the following has been added to subsection "3.3. Four-dimensional Communication.":

“Following the concept of four-dimensional communication, underground sensors or devices embedded in the soil can also be employed in order to collect data on various parameters like moisture levels and nutrient content. In both cases, the main research challenge is to mitigate the increased signal attenuation caused by soil or water layer interference on the transmission path. This in turn impacts network topology, as there is a tradeoff between the energy consumption and communication range of underground/underwater sensor nodes. Thus, in this context, new technologies such as the Internet of Underground Things (IoUT) [65], Wireless Underground Sensor Networks (WUSNs) [66], and Internet of Underwater Things (IoWTs) [67], are gradually advancing.”

Additionally, a discussion about ROI (return-on-investment) of the process of implementing new technologies in agriculture should be considered, especially bearing in mind the recent changes and price issues in the world food market and production.  

During the initial phases of its development, smart farming has indeed encountered problems, and in certain cases, the deployed systems failed to produce the anticipated outcomes. Therefore, the ROI of implementing new technologies in agriculture is a key factor that should be carefully considered.

However, it is also important to note that the ROI can vary depending on the availability and cost of the mix of technologies employed. As previously mentioned, this study aims to identify research paths for technologies whose first phase of standardization will be around 2030; therefore, it is risky to even estimate for these technologies their availability, rate of adaptation, and most importantly, their cost in the next decade.

However, based on the significant progress that has been made in recent years, ICT has demonstrated its value in the agricultural sector. The most reliable assessment of ICT's impact on agriculture can be found in the Global Smart Farming Market Forecast, which predicts the market will grow from USD 12.80 billion in 2021 to USD 33.69 billion in 2029 (ref. to [26]). Therefore, the current trend is that smart farming will continue to grow, which will lead to economically feasible technological solutions. 

The text at the end of Section 2.1 was revised and expanded to emphasize this point:

“Farmers can employ data-driven agriculture techniques for a variety of purposes, including strategic planning, real-time monitoring, performance evaluation, predictive forecasting, optimization, and effective event management. However, the widespread adoption of these techniques has been hampered by a number of obstacles, such as limited connectivity solutions and expensive and, in some cases, inaccurate data collection. Despite early obstacles, smart farming has made significant advances in recent years, with Information and Communication Technology (ICT) demonstrating its value in the agricultural sector [25]. Overall, adopting data-driven strategies has proven to increase agricultural productivity by increasing crop yields, reducing input costs, and minimizing crop loss. According to the Global Smart Farming Market Forecast, the market is anticipated to experience substantial growth, with projections estimating its value to rise from USD 12.80 billion in 2021 to USD 33.69 billion by 2029 [26], providing a reliable assessment of the expected impact of ICT on agriculture.”

Most of the references have a year of publication typed in bold font, but some of them (for example 32, 34, 39, 53, 73, etc.) are not. Authors should make the text uniform.

We would like to thank the reviewer for his meticulous examination of our paper. We noticed this as well, however it's due to the way Latex (Overleaf) generates the references for the specific template. The year of publication is bolded in the references labeled as articles, but not in the others (conference proceedings, miscellaneous, etc.). We believe that this is something that the journal's editing team can correct.

Also, minor revisions of the English language are required.

We carefully reviewed the manuscript and made any edits that were required.

Author Response File: Author Response.pdf

Reviewer 3 Report

This paper introduce scope and technologies for Agriculture 5.0. This paper  can be improved. The explanation about the relationship between technologies is needed. The scenarios for agriculture 5.0 are required. The some future research topics for agriculture 5.0 are also needed. 

Author Response

-Reviewer #3

This paper introduce scope and technologies for Agriculture 5.0. This paper can be improved. The explanation about the relationship between technologies is needed. The scenarios for agriculture 5.0 are required. The some future research topics for agriculture 5.0 are also needed.

To improve the paper and provide a more comprehensive explanation of the respective technologies and their application to agriculture 5.0, we have extensively revised large parts of the manuscript, mainly in the following sections:

Abstract, Introduction,

2.1. Evolution of Agriculture: From Traditional Farming to Agriculture 5.0,

3.1. Edge Computing and Pervasive Artificial Intelligence,

3.3. Four-dimensional Communication,

3.4. Quantum Sensing, Communication and Computation,

3.5. Terra-Hertz Communications,

4.1. Enhancing Sensing Accuracy and Reliability,

4.4. More Accurate and Reliable AI-based Applications,

4.5. Enhanced Remote Disease Assessment and Augmented Reality,

4.6. Enabling Secure and Transparent Transactions.

The new technologies expected to be developed and standardized for the 6G IoT infrastructure are extensively discussed in Section 3 titled "6G IoT: New Technologies and Emerging Services."

In the subsequent section, titled "4. The Future of Farming: 6G-IoT Applications in Agriculture 5.0," we explore various scenarios depicting how these new 6G IoT technologies are expected to be applied to Agriculture 5.0. Consequently, each subsection within Section 4 represents a future research path for Agriculture 5.0.

Author Response File: Author Response.pdf

Round 2

Reviewer 1 Report

No further questions. 

Reviewer 3 Report

The revised version has addressed my previous comments.