Smart supply chains go global: AI and IoT power predictive logistics

The report outlines how IoT devices, from RFID tags and GPS trackers to smart shelves and robotic pickers, are increasingly embedded in global supply chains to provide real-time data about product location, temperature, movement, and condition. Meanwhile, AI tools such as machine learning, deep learning, and reinforcement learning are being used to process the vast amounts of sensor data collected from these devices.

A new study has revealed how artificial intelligence (AI) and the Internet of Things (IoT) are fundamentally transforming supply chain operations, offering a detailed mapping of technological trends, challenges, and future trajectories across a range of industries. Published in the journal IoT by researchers at the University of Texas at Arlington, the comprehensive review analyzes over a decade of scientific literature using a dual approach: bibliometric analysis and topic modeling. It presents the most in-depth overview to date of how AI and IoT technologies intersect to drive the digitalization, resilience, and sustainability of modern supply chains.

The study, titled "Driving Supply Chain Transformation with IoT and AI Integration: A Dual Approach Using Bibliometric Analysis and Topic Modeling", examined 810 academic publications published between 2011 and 2024, identifying key thematic developments and rising technologies in supply chain digitization. It found that the convergence of AI and IoT has accelerated sharply since 2020, with significant research attention directed toward logistics optimization, cold-chain management, predictive maintenance, smart manufacturing, and supply chain resilience during disruptions like COVID-19. The authors observed a growing interest in advanced concepts such as federated learning, digital twins, edge computing, and blockchain, all of which are being incorporated into modern supply chain frameworks to address core operational pain points.

The report outlines how IoT devices, from RFID tags and GPS trackers to smart shelves and robotic pickers, are increasingly embedded in global supply chains to provide real-time data about product location, temperature, movement, and condition. Meanwhile, AI tools such as machine learning, deep learning, and reinforcement learning are being used to process the vast amounts of sensor data collected from these devices. The integration of these technologies has enabled real-time analytics, anomaly detection, and demand forecasting, improving transparency and decision-making at every stage of the chain.

Despite the potential, the researchers caution that implementing these technologies comes with considerable challenges. Key barriers include data security vulnerabilities, privacy concerns, interoperability issues, and a lack of standardization across platforms. The study notes that the heavy reliance on centralized cloud infrastructure exposes sensitive operational data to cyberattacks and compromises the integrity of systems managing critical supply functions. AI models, especially deep learning ones, often require significant computational power and are difficult to interpret, making them unsuitable for edge environments with limited resources or strict regulatory oversight.

To address these barriers, the authors identify federated learning as a promising solution. This decentralized training method allows machine learning models to be trained across multiple edge devices without transferring raw data to a central server, reducing both bandwidth consumption and privacy risks. The study finds that federated learning, along with edge computing, is gaining traction as an efficient and privacy-preserving strategy for deploying AI in supply chain environments. These approaches also support faster response times and greater resilience in low-connectivity or remote regions.

The analysis also tracks the emergence of Industry 5.0, a conceptual evolution from Industry 4.0, which emphasizes not just automation but human-machine collaboration, personalized production, and socially responsible operations. Technologies like AI, IoT, and blockchain are viewed as critical enablers of this next phase, especially in light of global policy shifts toward sustainability, domestic production, and economic protectionism. The report highlights how policies such as the U.S. CHIPS Act and the European Green Deal are steering supply chain strategy toward localization, resilience, and environmental accountability.

In terms of industry-specific applications, the study found extensive work in agri-food supply chains, where IoT sensors and AI models are used for traceability, inventory forecasting, and precision agriculture. In healthcare, federated learning is being employed to secure sensitive patient data while supporting predictive analytics for medicine distribution and pandemic response. In the energy sector, smart grids and AI-driven optimization algorithms are enabling efficient energy distribution, theft detection, and demand response. Cold-chain logistics, a critical component in food and pharmaceutical delivery, is also benefiting from AI-enabled monitoring systems that can track perishable goods in real time to ensure quality and reduce waste.

The authors also conducted a co-citation network and keyword co-occurrence analysis, revealing the seven dominant research clusters in the field: smart manufacturing, AI and cybersecurity, smart agriculture, healthcare logistics, smart grids, water resource management, and energy efficiency. These clusters underscore the breadth of AI and IoT applications and the growing interdependence between physical and digital supply chain components.

On the methodological side, the study adopted a mixed quantitative-qualitative design by combining bibliometric tools like VOSviewer and Biblioshiny with topic modeling using Non-Negative Matrix Factorization (NMF). This allowed the authors to map influential studies, track keyword trends over time, and identify gaps in current research. The dataset was curated using PRISMA screening on the Web of Science database, ensuring the inclusion of high-quality and relevant publications from across sectors.

The paper calls for further exploration of privacy-preserving AI methods, cross-sector interoperability protocols, and Explainable AI models that make machine learning outcomes more transparent to decision-makers. It also urges deeper collaboration between academia, government, and industry to align regulatory frameworks with emerging digital technologies in global supply chains.