Convergence of blockchain, IoT, and machine learning: exploring opportunities and challenges – a systematic review

Over the past 10 years, the combination of the Internet of Things (IoT), blockchain, and machine learning (ML) has brought about significant changes in various industries. This convergence has led to groundbreaking advances in industries, such as Industry 4.0, smart cities, smart tourism, and transportation systems. However, the extensive usage of IoT devices has also posed significant challenges in terms of scalability, security, privacy, and data management, due to the ever-growing number of interconnected devices. This ongoing proliferation of networked devices contributes to complexity, as each element has networking capabilities that enable the exchange and collection of information both locally and remotely over the Internet. As IoT systems communicate primarily over the Internet, their inherent vulnerability to a wide range of security risks becomes increasingly apparent [1]. This underscores the need for robust security measures to protect the integrity and confidentiality of data within these interconnected systems. To meet this requirement, it is necessary to deploy advanced algorithms, ML models, and data analysis tools to efficiently navigate and extract valuable information from an ever-changing data environment. Thus, the integration becomes imperative to guarantee the all-important aspects of data security, transparency, and immutability. Blockchain's decentralized and tamper-proof nature offers a solution to the security challenges inherent in traditional centralized systems, providing a secure foundation for the exchange and management of sensitive data within the rapidly expanding IoT ecosystem.

In the context of IoT applications, ML has evolved as a formidable tool for understanding models and facilitating real-time decision-making. Nonetheless, the security and privacy concerns linked to centralized ML frameworks have instigated an exploration of decentralized frameworks, with blockchain technology (BCT) assuming a critical role in this paradigm shift.

The incorporation of BCT enables not only the resolution of complex security and confidentiality challenges associated with automatic learning but also the improvement of the global performance of automatic learning algorithms by providing a reliable and secure database [2]. This mutually beneficial relationship between blockchain and ML has the potential to have a revolutionary impact on the global economy by introducing transformative changes in how data are handled, secured, and used to make strategic decisions. Aside from its capabilities in the resolution of security issues, blockchain establishes a sturdy framework that is invaluable for storing complete records of various activities and decisions in the field of IoT applications. This combined functionality enhances both data exchange efficiency and reliability while fostering continuous innovation in IoT. The confluence of blockchain and automatic learning strengthens not only the foundations of secure data practices but also propels a wave of innovation and development in the interconnected IoT environment.

As detailed in several reviews [2,3,4,5,6,7,8], the combination of IoT, blockchain, and ML technology offers considerable potential for improving algorithm performance and guaranteeing prediction accuracy. Indeed, data stored on a blockchain network is invulnerable to manipulation and error, reducing the risk of introducing errors or biases into ML models.

This study aims to comprehensively analyze the complex integration of IoT, blockchain, and ML technologies across various fields, exploring their potential applications and emphasizing the transformative power of their combined synergy. It also examines real-world scenarios and applications, revealing the complexities of integrating these technologies and illustrating how they can work together to assist industries, such as Industry 4.0, smart factories, smart tourism, and advanced transportation systems. Furthermore, it delves into the challenges associated with merging these disparate technologies, focusing on security, privacy, and data management issues. Simultaneously, the study identifies and explores the numerous opportunities presented by their seamless integration, offering a forward-looking perspective on potential innovations and improvements in a rapidly evolving technology landscape.

Compared to existing reviews, this study provides a valuable roadmap for understanding the opportunities and challenges related to IoT integration, blockchain, and ML. It combines a comprehensive review of existing research with real-world applications, including case studies from supply chain management, healthcare, smart cities, and renewable energy, demonstrating the practical impact of these technologies. This study also identifies crucial areas for future research, focusing on challenges such as data complexity, scalability, security, explainability, and regulatory frameworks, highlighting priorities for future development to fully realize the potential of these technologies.

To ensure a comprehensive and accurate analysis, an efficient review process has been implemented, as illustrated in Figure 1. Initially, a search was conducted using relevant keywords such as “AI,” “IoT,” and “blockchain,” which resulted in a collection of 130 articles. This initial pool of articles covered a broad spectrum of research within these fields.

Figure 1:

Research methodology.

The next step involved removing duplicate entries, which refined the set to 74 unique research papers. This crucial step ensured that each piece of research considered in the analysis was distinct and contributed uniquely to the understanding of the intersection of Artificial Intelligence (AI), IoT, and blockchain.

Subsequently, the abstracts of these 74 papers were meticulously reviewed. This review process was essential for determining the relevance of each article to the specific intersection of AI, IoT, and blockchain technologies. Articles that did not directly address the integration or application of these three technologies were excluded from further consideration. This careful curation resulted in a final selection of 64 research publications that were deemed most relevant for an in-depth review.

Figure 2 provides a detailed breakdown of the selected articles by topic, further highlighting the focus areas of the research. This systematic categorization of articles ensured a clear understanding of the specific domains within AI, IoT, and blockchain where significant research contributions have been made. For instance, some articles focused on the application of blockchain to enhance security in IoT networks, while others explored the use of AI for predictive maintenance in IoT devices.

Figure 2:

Elements percentile review. IoT, Internet of things.

This systematic approach not only ensured the inclusion of the most relevant and recent research but also facilitated a comprehensive understanding of how these technologies are being integrated across diverse fields. Through critical selection and organization of research, we provided a nuanced view of current advancements, identified emerging trends, and pinpointed areas for further exploration. This methodical process underscores the importance of a rigorous review in synthesizing a vast body of literature into coherent and insightful findings.

The survey is structured as follows. Section 2 provides an overview of existing survey articles, emphasizing the unique contributions and distinctions of this study. Section 3 introduces foundational concepts related to ML, IoT, and blockchain. Subsequent sections explore blockchain applications in IoT frameworks (Section 4), IoT and AI applications (Section 5), blockchain and AI integration (Section 6), and the combined applications of blockchain and AI with IoT architectures (Section 7). Section 8 identifies crucial research challenges for future exploration, while Section 9 concludes by summarizing the key findings and offering insights into the potential of this convergence for developing intelligent and secure IoT systems.

The overall structure of this study is illustrated in Figure 3.

Figure 3:

Study organization. AI, Artificial Intelligence; IoT, Internet of things.

To ensure clarity and consistency, Table 1 presents a comprehensive list of notations and their definitions used throughout this article.

Several reviews have been conducted to explore the integration of IoT, blockchain, and ML to enhance the security, scalability, and intelligence of IoT-based models. For example, a previous study [4] provides a comprehensive overview of various IoT, blockchain, and ML applications in the construction industry. The authors address challenges in productivity, time management, effectiveness, and quality and attribute these issues to limited awareness of contemporary technologies prevalent in other disciplines. Likewise, another study on the field of smart healthcare provides insights into the challenges and opportunities of integrating these technologies into healthcare [8]. Additionally, research [2] analyzes using BCT as a framework for storing records of various activities and decisions in the context of IoT. Furthermore, existing research discusses the integration of IoT with BCT and AI technologies and highlights the potential synergies and impacts of such integration [3]. Based on this context, Table 2 provides an overview of the related review works in several fields, offering a comprehensive examination on how the integration of IoT, blockchain, and AI has been explored and analyzed in existing literature.

Our review presents several distinctive strengths in comparison with existing works in the field of IoT, blockchain, and AI integration. While some research, such as the one on blockchain and AI integration in IoT [3], explores the potential benefits and challenges of this combination, our study goes further by providing a detailed analysis of specific synergies and proposing concrete solutions to overcome the identified challenges. Furthermore, while some studies (e.g., [9]) focus on security challenges using ML, AI, and BCT technologies, our contribution stands out by an in-depth analysis of standardized security protocols and advanced cryptographic methods.

Furthermore, our research differentiates itself by an in-depth analysis of interoperability aspects, proposing frameworks for smoother integration between these technologies, unlike a previous study [7] that focuses on overall technical challenges. Furthermore, our study proposes practical solutions to overcome the limitations of scalability and scalable blockchain architectures specific to IoT.

While the review [11] offers a comprehensive taxonomy and analysis of the challenges and opportunities in integrating blockchain and decentralized AI for cybersecurity, our study proposes concrete solutions for scalability and security challenges, as well as policy recommendations for responsible deployment of these technologies.

We do not limit ourselves to a single specific domain, but extend our analysis to various sectors. Unlike reviews focused solely on one sector such as logistics [13], construction [4], and colorectal cancer treatment [12], our research proposes integrated and standardized frameworks applicable to multiple industries. We identify specific challenges in scalability and data management, and propose solutions to overcome these obstacles.

Furthermore, we go beyond current reviews (e.g., [14]) by proposing policy recommendations and concrete solutions for scalability and security, integrating advanced cryptographic methods and data privacy frameworks. This approach ensures responsible and innovative development of IoT, blockchain, and AI technologies.

This survey is proposed based on existing relevant research to better understand the challenges and opportunities related to the integration of BCTs and AI in IoT systems.

Specific contributions to this study include:

We review related studies on the use of blockchain and AI in IoT. To highlight the improvements and enhancements achieved through the integration of these technologies, we divide the discussion into three principal areas: the integration of IoT with BCT, IoT with AI, AI with BCT, and the comprehensive integration of IoT with both blockchain and AI. This detailed examination allows us to understand the unique contributions and benefits of each type of integration.

The survey highlights the strengths and weaknesses of existing research on integrating IoT with blockchain and AI. Many crucial factors, such as decentralization, security, privacy, confidentiality, authentication, performance, and resilience, are examined. By analyzing these parameters, we aim to provide a balanced view of the advantages and disadvantages of the current approaches in order to identify areas that require further improvement.

Additionally, we identify various research challenges and future areas for designing effective IoT frameworks that integrate BCT and AI. This discussion aims to enable efficient IoT models with significantly improved application performance. We explore the obstacles, such as data complexity, scalability, and security, which need to be addressed, and suggest potential research directions to face these challenges and exploit the full potential of these technologies.

Key points of our study include a comprehensive review that synthesizes the current state of research on blockchain, AI, and IoT integration, providing a valuable resource for researchers and practitioners. We systematically categorize different types of integrations and highlight their unique benefits and challenges. By analyzing critical factors such as security, scalability, and interoperability, we identify gaps in existing research and propose future directions. These points provide a clear overview of the current landscape and guide future research efforts.

Our contributions are multifaceted. First, we provide an in-depth analysis of the strengths and weaknesses of existing research, providing a balanced view of current approaches. Second, we identify key research challenges and suggest potential solutions to address them, paving the way for more effective IoT frameworks. Third, we highlight the importance of developing standardized protocols, scalable blockchain architectures, and advanced cryptographic methods to improve data security and privacy. Finally, we propose practical applications and policy-related recommendations to ensure the responsible development and deployment of these technologies. Our study serves as a basis for future explorations and offers insights that could lead to significant advances in the field.

The flowchart shown in Figure 4 summarizes and illustrates the contribution of this work by highlighting the key stages of the analysis, starting with a review of existing studies and culminating in the identification of challenges and future directions for research.

Figure 4:

Flow diagram of the contributions. AI, Artificial Intelligence; IoT, Internet of things.

IoT is a network of physical devices, vehicles, structures, and other objects that are equipped with sensors, software, and communication capabilities to enable data gathering and sharing [22]. It has recently gained popularity because of its potential for advancement in important industries such as transportation, healthcare, and manufacturing. The key component of IoT is the smooth exchange of sensor data generated by IoT devices via connections to an IoT gateway. This gateway serves as a hub streamlining massive data transfer from several IoT devices [23]. This integrated infrastructure ensures quick and continuous data interchange and analysis, allowing businesses to increase operational efficiency, streamline decision-making processes, and stimulate innovation.

The data can also be sent to an edge device where it is analyzed locally, reducing the volume of data sent to the cloud and minimizing bandwidth consumption. Furthermore, IoT applications use ML algorithms to analyze vast amounts of data from sensors connected to the cloud, which results in real-time IoT dashboards and alerts that provide visibility into key performance indicators (KPIs), average time between failures, and other data. Even though IoT has many benefits, it also has drawbacks such as cost, complexity, and evolution.

To optimize network bandwidth and minimize data storage in the cloud, some IoT applications can pre-process data at the edge. This involves sending data to local devices for initial analysis, reducing the amount of data sent to the cloud. Meanwhile, in the cloud, powerful ML algorithms can analyze massive datasets from connected sensors. This creates real-time dashboards and alerts with valuable information, including KPIs, mean time between failure (MTBF) statistics, etc. Despite its benefits, implementing IoT solutions comes with challenges such as cost, scalability, and system complexity.

ML is an important field of AI that enables machines to perform specific tasks using learning methods. Over time, ML models have demonstrated extraordinary skills, outperforming humans in a variety of disciplines by utilizing previous experiences to complete assigned tasks [28].

3.3.1.

The ML approaches

ML algorithms are classified into several key categories, each representing a distinct approach to learning. Figure 6, a visual representation of these paradigms, illustrates their core principles and relationships:

Figure 6:

ML algorithms. ML, machine learning.

Supervised learning is a widely used ML paradigm where the algorithm is provided with labeled training data, and it learns to predict the output for new inputs [29]. Supervised learning techniques include classification (predicting discrete labels) and regression (predicting continuous values). Some common supervised learning algorithms support SVMs [30].

Semi-supervised learning bridges the gap between supervised and unsupervised learning by exploiting both labeled and unlabeled data. This approach is particularly useful when labeled data is scarce or expensive to acquire, as it allows models to learn from a larger amount of information while benefiting from the guidance of labeled examples.

Unsupervised learning involves discovering hidden patterns and structures in unlabeled data, often through methods like clustering. This is in contrast to the labeled data used in supervised learning [29].

Reinforcement learning is a learning paradigm where an algorithm learns by interacting with a dynamic environment and receiving feedback in the form of rewards or penalties. The goal is to learn an optimal sequence of actions to maximize the cumulative reward [31].

3.3.2.

ML processes

Generally, building a new ML model generally consists of two main stages: training and testing, which are required for tasks such as prediction, classification, and clustering on fresh datasets (Figure 7). Data are emerging as a significant asset in ML, playing a key role in both preprocessing and training models. Initially, the ML model is trained on a specific dataset, with the effectiveness of the ML classifier increasing as the training data size grows. Following training, the model's prediction accuracy is evaluated using a new dataset, leading to deployment if accuracy fulfills the criteria; otherwise, more training is performed. Recent improvements have seen DL emerge as a significant ML subgroup aiming to mimic human cognitive processes. Rooted in cognitive theories, DL structures are inspired by neural networks, with applications ranging from object identification to facial recognition and traffic flow prediction, demonstrating the range of its influence.

Figure 7:

Building ML model steps. ML, machine learning.

ML algorithms have proven useful in a variety of industries, including transportation, image processing, and the banking sector [32]. A broad range of models exist in the field of ML to handle various issue types, with SVM, artificial neural networks (ANN), and decision trees among the most commonly used [33].

The integration of ML and DL techniques with IoT is a promising direction to address the security challenges posed by the scale, heterogeneity, and real-time nature of IoT networks. Continued research in this area is crucial to develop robust, scalable, and privacy-preserving security solutions for the IoT ecosystem. Table 6 outlines several studies that focus on these aspects, providing insights into their challenges and proposed solutions.

The combination of BCT and AI presents a powerful synergy that offers numerous benefits across various industries. Blockchain's decentralized and immutable ledger provides a secure environment for storing and sharing data, which is crucial for AI systems that require vast amounts of reliable data to function effectively. This integration ensures data integrity, preventing tampering and unauthorized access, while also allowing for verifiable and traceable AI decision-making processes, fostering trust among stakeholders. Furthermore, blockchain can facilitate the sharing of high-quality, verified data across different entities, which can then be used by AI algorithms to generate more accurate and insightful predictions, particularly valuable in the fields like healthcare. The integration of blockchain and AI can also enable decentralized AI models, where multiple parties can contribute to and benefit from AI advancements without the need for a central authority, promoting innovation and collaboration. Finally, the combination of these technologies can enhance automation and efficiency in various processes, such as using smart contracts powered by blockchain to automate transactions and agreements based on AI-generated conditions and insights, reducing the need for intermediaries and accelerating operations across diverse applications.

The use of combined BCT and AI has proven their efficiency in several domains by improving security, efficiency, and transparency. This synergy is particularly impactful in domains like healthcare, where it enhances data management and transaction authentication, ensuring the integrity of patient records. In transportation, the integration of BCT and AI boosts operational trustworthiness and efficiency, providing innovative solutions to persistent challenges. These advancements demonstrate the significant potential of BCT and AI to address critical issues and elevate performance across various sectors described in Table 7:

The integration of blockchain and AI has the potential to revolutionize various industries by enhancing security, efficiency, and transparency. However, this integration is not without its challenges that must be addressed to ensure seamless operation. Primary challenges are given as follows.

• Data complexity and interoperability

Data complexity and interoperability are two major challenges in integrating blockchain and AI. BCT is based on decentralized and distributed data storage, which makes it difficult to integrate with AI systems that require structured and standardized data. This is because blockchain uses distributed data storage, which can make it difficult to access and manipulate data for AI systems that require organized and standardized data.

According to research detailed in a comprehensive review of the confluence of AI and blockchain technologies [54], the lack of interoperability between different blockchain platforms and AI systems can significantly hinder the smooth exchange of data. This review [54] has highlighted that the synergistic combination of these technologies, while beneficial in improving the performance and efficiency of existing information and communication technology (ICT) systems, is still in its infancy and subject to ongoing exploration.

Furthermore, the effectiveness of ML models depends on both the volume and quality of the data. According to a previous research [55], AI-based solutions to blockchain-related problems are often faced with limited or unreliable datasets. The lack of reference datasets that are free from mislabeling and have high diversity is evident.

The main data challenges in integrating blockchain and AI are as follows:

• ➢

  The complexity of data on blockchain, which is often unstructured and distributed, makes it difficult for AI systems to use.

• ➢

  The lack of interoperability between different blockchain platforms and AI systems hinders the smooth exchange of data.

• ➢

  The limited volume and quality of datasets are available to effectively train AI models, in particular the lack of high-quality reference datasets.

• Scalability and replicability

Another significant challenge is the scalability and performance of blockchain and AI systems. AI systems require substantial amounts of data to function effectively, making the scalability of blockchain platforms a critical issue. The current limitations of BCT in terms of transaction processing capacity and data storage can hinder the efficient integration of AI systems, leading to delays and inefficiencies. This problem has also been highlighted in several studies. One such study [55] details why scalability is considered a major issue in systems that combine blockchain and AI. According to the study, the limited transaction throughput and the high latency of current blockchain platforms can significantly impair the performance of AI applications that rely on real-time data processing. Additionally, the immense data storage requirements of AI models can overwhelm blockchain networks, causing further performance bottlenecks and reducing overall system efficiency.

Regarding reproducibility in ML, which is crucial for the credibility of results, a major challenge arises when applying it to blockchain. Ensuring consistent results is crucial to developing reliable AI-enhanced blockchain systems [55]. However, several factors hinder this achievement, including the scarcity of standardized datasets, rapid advances in AI and blockchain, and complex interactions between algorithms and blockchain structures. The lack of reproducibility can hinder the adoption of AI-based solutions in blockchain, making resolving this issue indispensable for the successful integration of AI and blockchain technologies.

• Security and privacy

One of the most important challenges in merging BCT and AI technologies lies in ensuring security and privacy [56]. This integration introduces complexities that necessitate careful attention to safeguarding sensitive data and preventing unauthorized access. According to a detailed examination of AI and blockchain convergence, the amalgamation of these technologies brings forth intricate security implications that demand thorough consideration.

AI systems rely heavily on extensive datasets for effective operation [56]. However, this reliance also amplifies the risks associated with data breaches and unauthorized access. The fusion of AI and blockchain exacerbates these concerns, as blockchain networks serve as repositories for sensitive transactional data that may become targets for malicious entities. This interconnectedness expands the potential attack surface, heightening the risk of data compromise.

Furthermore, leveraging AI within blockchain ecosystems introduces novel vulnerabilities [57]. Sophisticated AI algorithms can be exploited to launch intricate attacks on blockchain networks, utilizing their analytical capabilities to pinpoint and exploit system weaknesses. These AI-driven threats may target consensus mechanisms, manipulate transactional data, or circumvent security protocols, posing substantial risks to the integrity and stability of blockchain networks.

• Explainability and transparency

The integration of blockchain and AI also raises concerns about explainability and transparency. AI systems are often opaque and difficult to understand, which can make it challenging to identify and address biases and errors. The use of BCT can provide some transparency, but it is essential to ensure that the AI systems used in conjunction with blockchain are transparent and explainable.

• Regulatory frameworks

The integration of blockchain and AI also requires a clear regulatory framework. As AI systems are increasingly used in various industries, there is a need for regulatory bodies to develop guidelines and regulations that address the unique challenges and risks associated with AI-powered blockchain systems. In conclusion, the integration of blockchain and AI is a complex and challenging task. Addressing the data complexity, scalability, security, explainability, and regulatory challenges will be crucial to ensuring a seamless integration of these technologies. By understanding these challenges, we can work toward developing more efficient, secure, and transparent AI-powered blockchain systems that can revolutionize various industries.

Several studies have examined the combination of blockchain, IoT, and AI technologies and identified a range of transformative benefits [5, 58, 59]. This combination offers disruptive benefits such as increased security, efficiency, and intelligence in data management systems. A detailed overview of key benefits is listed in Table 8:

• Enhanced data security

The rapid adoption of IoT comes with inherent security risks, particularly due to the limited processing power and vulnerability of connected devices. BCT offers a promising solution by leveraging its decentralized and immutable ledger, which ensures the integrity of data generated by IoT devices [9]. This inherent security feature safeguards against unauthorized access and tampering, allowing AI algorithms to safely analyze sensitive data without compromising privacy [9]. This enhanced data security is crucial for building trust in IoT ecosystems, allowing for the safe and secure deployment of these technologies in diverse applications. As highlighted in Table 9, blockchain excels in securing data by offering features like data encryption, hashing, authentication, and integrity, which are often limited or absent in traditional IoT and AI systems.

• Improved decision-making

The convergence of AI and BCT revolutionizes decision-making within IoT. Blockchain's immutable ledger ensures data integrity and authenticity, providing a foundation of trust for AI algorithms. This enables AI to generate more reliable insights from real-time data streams, facilitating predictive analytics, anomaly detection, and personalized recommendations [60]. For instance, in smart transportation, AI can analyze traffic patterns, weather data, and vehicle sensor readings to optimize traffic flow and improve safety. Blockchain can ensure the integrity of these data sources, making the AI's insights more trustworthy and actionable, leading to better decision-making. This synergy enhances trust in IoT systems by guaranteeing data security, transparency, and accountability, leading to more informed, efficient, and trustworthy actions across diverse IoT applications.

• Increased efficiency

The automation capabilities of AI, when fueled by real-time data from IoT devices and secured by blockchain, significantly enhance operational efficiency. AI can automate routine tasks, predict maintenance needs, and optimize processes, reducing manual intervention and accelerating response times. Blockchain ensures that the data driving these automations is accurate and tamper-proof [60]. For example, in a manufacturing setting, AI could analyzesensor data from machines to predict potential failures, allowing for proactive maintenance and minimizing downtime. Blockchain could then provide an immutable record of these predictions and maintenance actions, ensuring transparency and accountability in the process. This combination of AI and blockchain empowers organizations to streamline operations, optimize resource allocation, and improve overall productivity within their IoT ecosystems.

• Traceability

The convergence of blockchain, IoT, and AI creates a powerful ecosystem that ensures unparalleled traceability within complex systems, transforming industries where data integrity and transparency are critical. By seamlessly integrating these technologies, we gain the ability to track the origin, movement, and transformation of data across IoT networks with unprecedented accuracy and reliability [7].

In supply chain management [7], sensors embedded within goods constantly transmit data about their location, condition, and handling to a blockchain-based platform. AI algorithms can then analyze these data in real time, identifying potential risks or anomalies and triggering preventative actions. Blockchain ensures that every data point is immutably recorded, creating a transparent and auditable trail that can be accessed by all stakeholders. This eliminates the possibility of tampering or manipulation, fostering trust and accountability within the supply chain.

Similarly, in healthcare [7], this convergence can revolutionize patient data management. Sensors embedded within medical devices can constantly transmit patient vitals and medical information to a secure blockchain platform. AI algorithms can analyze these data, identifying potential health risks and suggesting personalized treatment plans. Blockchain guarantees the immutability and security of these data, enabling secure and transparent data sharing between healthcare providers while safeguarding patient privacy.

This powerful combination of blockchain, IoT, and AI not only ensures traceability but also empowers stakeholders with actionable insights, enabling them to make more informed and efficient decisions. This transformative approach promises to reshape diverse industries, from supply chain management to healthcare, fostering greater trust, transparency, and innovation within the IoT ecosystem.

• Cost reduction

The combined use of AI, IoT, and blockchain can lead to significant cost reductions. AI and IoT optimize operational processes, reducing waste and inefficiencies. Blockchain reduces the need for intermediaries by providing a secure and transparent system for transactions and data exchanges. Together, these technologies lower operational costs and enhance the overall efficiency of business processes [7].

We can also streamline this combination as shown in Figure 8.

• ➢

  IoT data generation: The IoT devices are the data sources, which constantly collect information from the real world (e.g., sensors monitoring energy usage, weather conditions, or even traffic patterns).

• ➢

  Blockchain data security and integrity: Blockchain acts as a secure and tamper-proof ledger for the data generated by the IoT devices. It ensures that the data are trustworthy, accurate, and protected from manipulation.

• ➢

  AI data analysis and decision: AI algorithms analyze the secure and reliable data stored on the blockchain. This analysis can lead to valuable insights and predictions, allowing for informed decision-making.

• ➢

  Real-time data processing: The system emphasizes real-time data processing, enabling swift responses and proactive adjustments based on the insights gleaned from the data.

Figure 8:

Combination of IoT, BC, and AI processes. AI, Artificial Intelligence; IoT, Internet of things.

Several sectors are transforming because of the convergence of blockchain, IoT, and AI, which is opening new opportunities for accountability, efficiency, and transparency. The transforming potential of this combination is demonstrated by the fascinating case studies and examples of applications detailed in Table 10.

• Supply chain management

Traceability and counterfeiting: Blockchain's immutable ledger provides a tamper-proof record of the journey of goods from their origin to the end consumer. IoT sensors embedded in products constantly transmit data about their location, condition, and handling to the blockchain, enabling real-time tracking and verification. AI algorithms can analyze these data, identifying potential risks or anomalies and triggering preventative actions. This creates a highly secure and transparent system that can prevent counterfeiting, ensure product quality, and enhance consumer trust. Table 11 gives a detailed overview of the contributions of IoT, blockchain, and AI across various supply chains and industrial applications.

• Healthcare

Secure data sharing and privacy: Blockchain offers a secure and transparent platform for storing and sharing patient data, protecting individual privacy while enabling efficient data access for research and treatment. IoT sensors embedded in medical devices can continuously monitor patient health, transmitting data to a blockchain-based platform. AI algorithms can then analyze these data to identify potential health risks, personalize treatment plans, and improve patient outcomes. An in-depth overview at how blockchain, IoT, and AI address various challenges is given in Table 12.

• Smart cities

Resource management and optimization: Blockchain, IoT, and AI can work together to optimize resource management and improve efficiency in smart cities. Sensors embedded in traffic lights, waste bins, and energy grids can transmit data to a blockchain-based platform. AI algorithms can analyze these data to optimize traffic flow, predict waste collection needs, and manage energy consumption more efficiently. This creates a more sustainable and efficient urban environment. A comprehensive summary is given in Table 13.

• Renewable energy

Tracking and transparency: Blockchain can be used to track the origin and flow of renewable energy, ensuring transparency and incentivizing its production and consumption. IoT sensors can monitor the output of solar panels, wind turbines, and other renewable energy sources. AI algorithms can analyze these data to optimize energy production and distribution, ensuring greater efficiency and maximizing the use of renewable resources.

The integration of blockchain and AI in energy cloud management (ECM) is a crucial step toward a more sustainable and efficient energy future, particularly for facilitating the widespread adoption of renewable energy sources. Research (e.g., [70]) highlights the significant role of this integration. It underscores the potential of blockchain to create a secure and transparent environment for P2P energy trading, where prosumers can sell excess energy from their renewable sources (e.g., solar and wind) directly to neighbors, incentivizing renewable energy adoption. AI algorithms, in turn, can optimize energy production and distribution by accurately predicting demand from renewable sources and facilitating better integration into the grid. This research also emphasizes the need for robust security measures to protect data in blockchain-based energy systems, addressing concerns about data breaches and ensuring the trust and privacy of consumers. Overall, this research contributes to the development of a more secure, transparent, and efficient energy ecosystem that encourages the greater use of renewable energy sources.