However, supply chain management (SCM) has never been an incident to the success of the industry/commerce; in the modern dynamic business environment, it is necessary than ever and more complex than it has been at any point. The necessity to possess dynamic and smart supply chain systems has considerably been boosted by the forces of globalization, rivalry, and alterations in the demands of the consumers. In the vast majority of cases, the traditional techniques of efficiently producing the supply chain given predominantly historical information and in the deterministic forecast models do not adapt to unstable markets, geopolitical vulnerabilities, and unpredictable epidermis like pandemic or natural disasters. These limitations have made firms create more advanced, data-driven and intelligent methods of how they should operate their supply chains. The most recent approach to the challenge is predictive analytics is driven by Artificial Intelligence (AI) because this method can enable an organization to predict interruptions, maximize resources, and make preemptive decisions with more accuracy [9].

The rationale of conducting this research is that it is quite becoming evident that the supply chains in the modern world are no more a linear system that can be predicted but they are dynamic networks within which uncertainties can produce ripple impacts across the production, distribution and consumption networks. Using the case of a late supply of raw materials in one part of the globe, production in the other will be halted causing ripple effects in the magnitude of inventory of a given supply, transportation schedule-- customer satisfaction [15]. In that regard, predictive analytics provided by AI will become a strategic facilitator, since it will be able to allow organizations to transform vast quantities of both structured and unstructured information into actionable intelligence. Machine learning algorithms, natural language processing, and deep learning technologies are applicable in predictive analytics to identify concealed trends in data, foretell the future, admission supply chain services in real-time, as well as optimism.

The alternative reason why it is crucial that this should be done is that AI has the potential of improving the resiliency of supply chains. The shock felt by the changes in the world over the past couple of years has proved that the supply chains are susceptible to the approaching shock, which can be a health pandemic, an armed conflict of a geopolitical nature, or a disaster caused by climate changes. The classical forecasting models that exclusively rely on values of the history struggle to justify such a spectacular happening. Rather AI models can be continuously trained on their own with numerous data as sources provided by IoT sensors, social media, market dynamics, and supplier risk data. This flexibility is not only capable of helping organizations predict demand but also can assist the organization highlight their vulnerability in order to institute contingency plans to enhance organizational resilience and competitiveness [13].

The research assumes three large objectives. Firstly, it will introduce a detailed examination of how AI-based predictive analytics can be applied in some of the supply chain management processes like demand prediction, inventory optimization, logistics optimization, and supply risk management. Second, the purpose of judging the predictive analytics effectiveness is to determine it via the supply chain through efficiency and reduction of costs and correct decision making. Third, the study aims at highlighting such pragmatic concerns as data quality, implementation cost, and the workforce preparedness and recommending the further studies and practice. With these objectives, the work will address this gap in the developments of AI theory and its subsequent applications in supply chains in the world.

In brief, the adoption of AI-predictive analytics in the engagement to run any supply chain is not only an extra pass of improving things but also the transformation of the paradigm. It gives firms flexibility to shift to being reactive to making proactive and even prescriptive decisions. The paper, therefore, targets to unearth opportunities, constraints, and implications of AI implementation in supply chains, but in a broader sense, this is made possible in enabling the development of adaptive, efficient and sustainable supply chain ecosystem [10].

Novelty and Contribution

The originality of this piece of work is in viewing AI-accomplished predictive analytics as a technological revolution and an enabling aspect of supply chain management in general. Despite the fact that current literature has consolidated this problem on a very narrow level of demand prediction or logistical plans, the current paper brings a multi-dimensional project that integrates predictive analytics within different supply chain aspects, including inventory, logistics and reduction of the supplier risk. This study seems to be in conflict with traditional supply chain research studies in that they expected the research to concentrate on the future concentrated research of the studies and not on analyzing past the way more traditional research had gone [14].

Among the most significant contributions made to work is the outline of the primary restrictions regarding practical implementation, not to mention that the proposed solutions are proposed. The paper does not show AI as an ideal solution but mentions some critical challenges, such as the high price of a solution, data integrity issue, the risk of hacking, not to mention the cultural resistance of the organization. That is why it provides a real-life foundation to the decision-makers in order to experiment with the AI practise within their supply chains. Moreover, explainable AI as another step in achieving transparency and blockchain were mentioned in the research as one of the future directions alongside the organization presented in real-time that will enable various stakeholders to share knowledge and coordinate their activities [7].

The emphasis on the principle of resilience and sustainability is another important input. Besides effectiveness in work and cost reduction, the implementation of artificial intelligence in predictive analytics has also been discussed as fundamental in the development of the supply chain resistant to disruptions and contributing to environmental sustainability due to the reduction of wastes and optimization of traffic paths, as well as minimization of unused stocks, in this work. This way, not only are the work positions implying that predictive analytics could serve as an assurance of efficiency, but are laying a groundwork according to which robust and sustainable supply chains may be established.

Overall, the novelty and value of this paper is interpreted in the fact that values possesses a balanced view recognizes negotiation - neither unique to the theoretical developments within the field of AI nor overlooks the issues currently lying in its implementation, but rather incorporates a vision of the future based on the provision of adaptability, transparency, and sustainability to the future or supply chain management [11].

Related Works

In 2024 S. B. Dhal et.al. and D. Kar et.al., [12] introduced the appearance of the Artificial Intelligence (AI) application to supply chain management is a topic that has been widely discussed in the past few years, in part due to its ability to work with large amounts of data, and resulting into practical solutions, which would not have readily been sought out through other conventional approaches. The previous analyses have shown the consistency in the methods through which predictive analytics can introduce a major change to the effectiveness of the demand forecasts by discovering the tendencies in which the consumer buys the commodity and readjust its prediction dynamically in this aspect. It has been noted that machine learning models, particularly models based on time-series data and recurrent neural networks are more effective in the short run and long run prediction as compared to the traditional statistical approaches. Such innovations are needed, since sharp prediction is the key to the effective inventory management and logistics policy. Predictive analytics will enable them to be better placed to diminish their stock outs, and buy excesses to achieve their higher degrees of efficiency and customer satisfaction.

In the environment of the control of stock, the literature demonstrates the suitability of the application of predictive analytics to optimize the level of stock security and reorder points. The traditional model of inventories has been previously operating with constant demand and predetermined set lead times that barely reflect the actual conditions of the global supply chain that do not remain constant throughout the time [16]. Contrary predictive models can utilize different kinds of data such as the history of supplier’s performance, weather pattern and market situation of the region that can lead to a more precise forecasting. Evidence provided in this case study suggests that predictive analytics tool, which applies the use of AI, will reduce the price of inventory holding by making the dynamical alteration through real-time optimization. These findings point out that predictive strategies do not merely transform the process of structuring costs in a more efficient way, as they also increase the level of services where their ability to get the product ready at the opportune moment is very crucial in these industries as the retail industry, the healthcare industry and the manufacturing industry.

In 2024 K. Kumar et.al. and A. Kumar et.al., [6] suggested the optimization of logistics and transport is also one more aspect that is being successfully developed. In literature that corresponds, predictive analytics are applicable in better fleet management and route planning in the case of reinforcement learning and/or optimization algorithms. To minimize delays, the predictive models are designed to help them maintain low costs of operation since they forecast the traffic and fuel consumption, delivery time, and the external elements such as weather interruptions. These have proved effective in high volume logistic networks to cause speed, dependability as well as sustainability. In addition, the heterogeneous models support the modelling of the transportation system at multi-modes, through which they can effectively and efficiently deliver products. The latter observations are the pointers to a broader shift to smartlogistical instruments that are actively based on the environmental changes and customer interests at every moment of time.

Supplier relationship management is one more significant field of study within the recent past. The supply chains are being prone to disruptions which can either be occasioned by the wavering resolve by the supplying suppliers, geopolitical conflict or the inaccessibility of raw materials. Supplier risk tracking using predictive analytics has been done on the basis of financial soundness, historical delivery and market indicators of suppliers. It has been proved by an analysis that predictive models can forecast possible fires of suppliers that may occur even before they do so therefore leaving the business with plentiful time to seek alternative supply sources or deflect production timelines. It is an active form of supplier management compared to the reactive operations system in the process of developing successful supplier relationship. By housekeeping its charitable prognostication capabilities into the analysis of its vendors, an enterprise will manage to promote its resistance significantly and alleviate the risk of a costly failure [5].

Emerging research has also explored the possibility of having AI-powered predictive analytics in conjunction with Internet of Things (IoT) in relation to big data platforms. One aspect can be discussed regarding Iot -ready sensors of warehouses, manufacturing houses, and transport vehicles that generate constant flows of real-time information. It has been used to train predictive models based on such data to monitor the health of an equipment making predictions of potential failure and permitting predictive maintenance policies. The methods reduce the down time and life span of crucial assets. Another discovery, which has been made is that 1:1 integration of both the IoT and predictive analytics leads to better transparency across the supply chain as stakeholders can now identify the product arrangements of the supply base and administrations of the suppliers to determine the goods they are moving and their existing location in terms of availability within their environments. With the help of such integration, the cooperation and trust of relationships with partners are enhanced, and the establishment of more coherent supply chain ecosystem is preconditioned.

Linear works have seen a major prevalence of the appearance of sustainability and environment factor. The carbon emitted has been reduced with the application of predictive analytics by choosing optimization in a transportation route and load size. In addition, forecasting statistics have been implemented on product returns that have enabled companies to prepare reverse logistics systems that reduce wastes and promote recycling. The papers reveal that, predictive analytics will result not only in efficiency and profitability, but in more abstract corporate social responsibility and sustainability targets. With the provision of the data-driven decisions, the organizations would be able to modify their supply chain plans to appear according to the environment and social objectives, benefiting both the company and the society.

Despite these developments being done, therefore, the associated studies are always described to be faced with numerous challenges. The poor state of data is also a major bane as predictive models can never be more than the data with which they are used. The imprecise, outdated and biased information can lead to the wrongful choice and faulty forecasting. Moreover, it is demonstrated in many sources that Aldo AI technologies are expensive to apply and are forbidden to small and medium companies. The other problem, which however, is likely to be experienced is integration with the current legacy systems and this may require a massive reorganization of the IT infrastructure. Finally, there is the issue of explain ability of predictive models, which are becoming increasingly popular where the concerned parties are asking questions on how the decisions are actually created. All these weaknesses necessitate such constant research on explainable AI and cost-effective methods of implementation.

In 2025 S. Al-Hourani et.al. and D. Weraikat et.al., [17] proposed the literature of closely related researches highly supports the transformational abilities of AI based predictive analytics in Supply Chain Management. The measures of demand anticipations, inventory control, performance, and removal of supplier and environmental sustainability risks are some of the common benefits reported in the literature. At the same time, it also releases realistic barriers that must be overcome in order to be utilized by numerous people. This two-sidedness leading into openness and limitation is the main pillar of the also investigation in this paper. Based on the former aspects of knowledge and bridging the gaps that exist in sources of literature, the study to be undertaken in this paper will be useful in the criminalization of the prediction analytics use that is aimed at creating resilient, adaptive, and sustainable chain of supply

The specified example of an AI-powered predictive analytics that will streamline the supply chain management operation could be regarded as a multi-phase system, not merely with respect to varying types of data integration, but also in the context of leveraging the latest approaches to machine learning models and supply chain optimization to formulate recommendations which should be undertaken during the decision-making process. Another concept on the methodology is flexibility, portability and real life in the industries that require efficacy in the supply chain processes. By merging both the structured and unstructured information based on predictive and prescriptive analytics, the proposed system will remove the issue of unpredictability in demand and inventory wastes, logistical failures and supplier risks. Consideration of modules in an inherited manner gives the methodology conceptualization the impact of creating an intelligent and adaptive system of decision making by the supply chain managers.

The initial phase of the methodology will entail data collection and preprocessing. Supply chains bring into existence a large amount of heterogeneous data in various forms like ERP systems, IoT-enabled sensors, and a history of customer and supplier performance and external data such as market trends and weather conditions. Raw data is however usually incomplete, irregular, or noisy which curtails its research. Hence, data intake is influential in determining that the feeds into predictive models are precise and sound. It is the phase of cleaning relatively missing values, eliminating redundancy, normalizing numerical data, encoding categorical data and combining data found in structured and unstructured form. Also, the use of the natural language processing methodology is utilized to get insights out of the textual information, including supplier reviews, mentions in social media, and customer feedback [2].

The second step concerns predictive modeling with machine learning algorithms and deep learning algorithms. Supply chain data of the past are utilized to train prediction models, which can resolve the time effect and multi-dimensional nonlinear changes. Some of the demand forecasting algorithms are regression models, random forests and recurrent neural networks (RNNs) [17]. The time-series models are most appropriate in the urge to rise the models and seasonal changes in demand. Similarly, classification algorithms are used to forecast the supplier risk and identify suppliers who can or cannot be able to deliver on time. Reinforcement learning strategies are also discussed to bring the dynamic logistics so as to get the optimal routing decisions which are subject to the changing conditions of the surrounding environment, such as the delay of route through traffic, weather and other abnormalities and so on.

The third stage deploys some layer of optimization which includes information which is currently in use to insert the foreseen ramifications to realistic actual decisions. In order to provide a balance between the levels of services and holding costs, as an example, following the development of a demand forecast, the optimization module forms the appropriate inventory which it should hold. Similarly, the predictive data regarding logistics is utilized to strategize the best transportation pattern, fleet transportation, and load planning strategy. Mathematic anticipating methods which are employed in such a layer are; linear programming, mixed integer programming and the heuristic approach used in making cost-effective decisions that will give an assurance that they are practical. The combination of optimization and predictive analytics will ensure that the insights lead to the creation of something tangible regarding the supply chain performance. The comprehensive performance attributes and metrics provide a balanced scorecard for evaluating supply chain health, allowing for precise measurement and targeted improvement initiatives through metric decomposition [8].

The fourth-stage is devoted to the assessment of the predictive models and optimization strategies and the validation. The models are robust as this is achieved by the insertion of the performance metrics of Mean Absolute Percentage Error (MAPE) demand forecasting, precision-recolor scores supplier risk prediction, as well as the percent cost savings during logistics optimization. Cross-validation is applied to determine the generalizability, whereas simulation environment is invoiced to estimate the performance when faced with various situations of disruption. An example is that demand shock simulation checks the degree to which the system can respond to immediate changes in demand whether they are upward or downward. Not only are they validated in historical datasets but also in historical control pilot programs in the real-life environment in a supply chain execution.

System integration and deployment comes in the fifth stage. The AI models get embedded into the already existing enterprise platform like ERP software and/or supply chain management software, through the APIs and cloud services. A dashboard interface will be developed to show predictive insights displayed in a form that can be understood easily by a supply chain manager to make a profound decision in a brief period. Some of the important points raised by the dashboard include accuracy of its forecast, inventory position, and transportation delays, as well as risks at a supplier. Notably, it is a scalable system that is capable of being used as a small scale system in individual organizations up to a large scale setup running global supply chains. They include security solutions and encryption to prevent cybercrime with respect to confidential data in the supply chain [3].

The last and sixth stage focuses on constant learning and improvement. The supply chains are very dynamic and hence the predictive model should always be updated with new data to ensure continuity in precision. Hence, it is a combination of online learning methodologies in which models are constantly modified as new data flows in through the sensors of the IoT, transactions and market external information flow. Opinions are linked to prediction systems thus creating a loop between predictive systems and the outcomes of any decision, which allows the models to develop according to shifting conditions. This step ensures that the way the methodology is designed is never fixed to anything singular but is flexible and upgrades as time passes to meet new challenges thus enabling the supply chain to become smarter and increasingly more resilient.

The flowchart below shows the end-to-end approach, that has the first step as gathering raw information, then the next process with preprocessing and predictive modeling with AI, and finally, the actionable information on optimizing forecasting, inventory management as well as the logistics program.

FIG. 1: PROPOSED AI-POWERED PREDICTIVE ANALYTICS FRAMEWORK FOR SUPPLY CHAIN OPTIMIZATION

The introduction of AI-based predictive analytics into the supply chain structure yielded substantial returns in the various spheres of operation, and shows quantifiable returns as compared to previous methods of operation. Among the most outstanding discoveries was in the field of demand forecasting. The conventional techniques were prone to seasonality, market variations, and inconsistent consumption patterns as a result of which, the actual inventory planning had inaccuracies and inefficiencies. As deep learning models and machine learning were used, the accuracy of the forecasts on demand increased greatly. As Figure 2: Demand Forecasting Accuracy Comparison demonstrated, the traditional forecasting accuracy was 72% with machine learning improving the accuracy to 85% and deep learning bringing the accuracy up to 90%. This gain was directly translated into increased planning, lowering of stockouts, and decreased surplus stock and increased the overall stability of the operations.

FIG. 2: DEMAND FORECASTING ACCURACY COMPARISON

The only other important discovery was in inventory management. The traditional processes are generally based on a fixed reorder point and a non-moving safety stock computation which in most cases cannot reflect the fast changing demand and supply characteristics. The predictive analytics processed with real time information saved much inventory costs. Figure 3: Inventory Cost Reduction Across Methods can witness this possibility, as it appears that traditional systems have decreased the cost only by approximately 5 per cent, the AI-advanced predictive analytics reduced it by 18 per cent. The accuracy of reorder point that predictive models could manage and optimize gave this a significant boost particularly with the safety stock optimization of safety inventories. These saving of holding costs are essential especially in an industry that has high storage often expensive or perishable goods where unnecessary holding of the inventory means losing.

FIG. 3: INVENTORY COST REDUCTION ACROSS METHODS

The other angle, whereby predictive analytics proved quite handy, was the optimization of logistics. In the case of the use of AI-based optimization models, transportation data analysis of numerous curves enabled the suggestion of a more effective pattern of routes and threats of weather factors. Figure 4: Logistics Delay Reduction The comparison indicates slight difference between the traditional methods of 7% reduction in the delays reduction whereas the optimized AI systems have indicated a reduction of 20%. This has not only made the mode of transport cheap but also made the customers to be happier as long as their packages reached the destination in time. It is important to note that these advantages were in line with the objectives of sustainability as the smoothness of the logistics routes reduced the quantities of unnecessary fuel consumed and carbon emissions.

FIG. 4: LOGISTICS DELAY REDUCTION

Along with the graphical representations, in Table 1: Performance measurements Comparison comparative data is also provided. The table elaborately conveys support to the high-grade AI enhanced predictive analytics in many ways. It was found that this increased the accuracy of the forecasting to 90 percent compared to 72 percent, inventory costs were lowered to 18 percent compared to 5 percent and also the logistical delay was enhanced to 20 percent compared to 7 percent. These findings validate the fact that predictive analytics presents both low-level and high benefits in comparison to conventional supply chain management activities.

TABLE 1: PERFORMANCE METRICS COMPARISON

In addition to quantitative performance the predictive analytics integration also improved qualitative aspects of supply chain management such as customer satisfaction, supplier risk management and improvement of service level. The AI-based systems delivered huge gains in the following areas as demonstrated in Table 2: Gains in Operational Efficiency. An example would be customer satisfaction which under the traditional models was constrained by delay and stockouts but under the AI governed models was extremely high. Supplier risk prediction that had a medium accuracy level in older systems developed tremendously with risk prediction model that was able to anticipate risks proactively. Similarly, the overall service quality level was also elevated concerning analytics created by AI, that allowed the companies to question its customers the minimum number of questions and construct long-term trust between the customers and the stakeholders.

TABLE 2: OPERATIONAL EFFICIENCY GAINS

Such findings confirm the stance that the use of AI-based predictive analytics can be called not only a technological advancement but also an ability to transform the Strategy that can transform the Supply chain management. Since the quantitative deliver major benefits regarding cost and efficiency, the qualitative ones present more resilience, better risk management and, enhanced customer involvement and interaction. It must be noted that these instead are conditional results and are tied on the effectiveness of the input data, good infrastructure, and organizational readiness to adopt new technologies. Lack of all these requirements can prevent a full increase in the potential benefits of predictive analytics [4]

AI-based predictive analytics can prove to be of great value in supply chain management with better forecasting, inventory optimization, optimization of logistics, and supplier risk management. It is an essential tool to a contemporary supply chain due to its capability to combine various data and produce solutions.

Nevertheless, the constraints in practice still exist. The prohibitive adoption expenses, reliance on clean and comprehensive datasets, cyber vulnerability and the difficulty of establishing AI within legacy systems are still limiting to many organizations. Moreover, some AI models are black-box and, as such, they pose a threat to trust and explain ability in the decisions taken.

Future trends must focus on explainable AI (XAI) as a way to be transparent, real-time internet of things (IoT) and blockchain as secure and adaptable supply chains, and human-AI collaborative systems that enhance the human decision-making process and add value instead of replace it. More than merely overcoming these obstacles, predictive analytics could become one of the foundations of resilient, sustainable, and globally responsive supply chains

1. Shamsuddoha, E. A. Khan, M. M. H. Chowdhury, and T. Nasir, “Revolutionizing supply chains: unleashing the power of AI-Driven intelligent automation and Real-Time information flow,” Information, vol. 16, no. 1, p. 26, Jan. 2025, doi: 10.3390/info16010026.
2. B. Dhal and D. Kar, “Transforming Agricultural Productivity with AI-Driven Forecasting: Innovations in Food Security and Supply Chain Optimization,” Forecasting, vol. 6, no. 4, pp. 925–951, Oct. 2024, doi: 10.3390/forecast6040046.
3. Daios, N. Kladovasilakis, A. Kelemis, and I. Kostavelis, “AI Applications in Supply Chain Management: A survey,” Applied Sciences, vol. 15, no. 5, p. 2775, Mar. 2025, doi: 10.3390/app15052775.
4. K. Vadaga, U. R. Dokuburra, H. Nekkanti, S. S. Gudla, and R. K. Kumari, “Digital transformation in Pharmaceuticals: The impact of AI on supply chain management,” Intelligent Hospital, p. 100008, Jul. 2025, doi: 10.1016/j.inhs.2025.100008.
5. Zaman, A. Shoomal, M. Jahanbakht, and D. Ozay, “Driving Supply Chain Transformation with IoT and AI Integration: A Dual Approach Using Bibliometric Analysis and Topic Modeling,” IoT, vol. 6, no. 2, p. 21, Mar. 2025, doi: 10.3390/iot6020021.
6. Kumar and A. Kumar, “Application of Optimization Models in Sustainable Supply Chain Management: A Systematic review based on PRISMA guidelines,” Process Integration and Optimization for Sustainability, Oct. 2024, doi: 10.1007/s41660-024-00455-x.
7. J. Serrano-Torres, A. L. López-Naranjo, P. L. Larrea-Cuadrado, and G. Mazón-Fierro, “Transformation of the dairy Supply Chain through Artificial intelligence: A Systematic review,” Sustainability, vol. 17, no. 3, p. 982, Jan. 2025, doi: 10.3390/su17030982.
8. Anumula, Sathish Krishna. "Next-Gen Supply Chains: A product Lifecycle Management Based Approach To Resilient And Sustainable Operations." International Journal of Managing Value and Supply Chains (IJMVSC) Vol.16, No.3, September 2025
9. A. Thani, M. Hadid, R. Padmanabhan, L. Kerbache, and A. Elomri, “Smart Food Supply Chain Management: A Bibliometric and Systematic Review,” Food and Humanity, p. 100736, Jul. 2025, doi: 10.1016/j.foohum.2025.100736.
10. Kayvanfar, A. Elomri, L. Kerbache, H. R. Vandchali, and A. E. Omri, “A review of decision support systems in the internet of things and supply chain and logistics using web content mining,” Supply Chain Analytics, vol. 6, p. 100063, Mar. 2024, doi: 10.1016/j.sca.2024.100063.
11. Shadkam and E. Irannezhad, “A comprehensive review of simulation optimization methods in agricultural supply chains and transition towards an agent-based intelligent digital framework for agriculture 4.0,” Engineering Applications of Artificial Intelligence, vol. 143, p. 109930, Jan. 2025, doi: 10.1016/j.engappai.2024.109930.
12. Madani, A. Saihi, and A. Abdelfatah, “A Systematic Review of Sustainable Supply Chain Network Design: Optimization Approaches and research Trends,” Sustainability, vol. 16, no. 8, p. 3226, Apr. 2024, doi: 10.3390/su16083226.
13. J. De Santana Filho, A. D. Paço, and P. D. Gaspar, “Artificial Intelligence and MCDA in Circular Economy: Governance strategies and optimization for reverse supply chains of solid waste,” Applied Sciences, vol. 15, no. 9, p. 4758, Apr. 2025, doi: 10.3390/app15094758.
14. Ramezani, M. Khazaei, F. Gholian-Jouybari, A. Sandoval-Correa, H. Bonakdari, and M. Hajiaghaei-Keshteli, “Turquoise hydrogen and waste optimization: A Bi-objective closed-loop and sustainable supply chain model for a case in Mexico,” Renewable and Sustainable Energy Reviews, vol. 195, p. 114329, Feb. 2024, doi: 10.1016/j.rser.2024.114329.
15. Yadav, R. K. Garg, and A. Sachdeva, “Artificial intelligence applications for information management in sustainable supply chain management: A systematic review and future research agenda,” International Journal of Information Management Data Insights, vol. 4, no. 2, p. 100292, Sep. 2024, doi: 10.1016/j.jjimei.2024.100292.
16. Waldhans, A. Albrecht, R. Ibald, D. Wollenweber, S.-J. Sy, and J. Kreyenschmidt, “Temperature control and Data Exchange in food Supply Chains: Current situation and the applicability of a digitalized System of Time–Temperature-Indicators to optimize temperature monitoring in different cold chains,” Journal of Packaging Technology and Research, vol. 8, no. 1, pp. 79–93, Feb. 2024, doi: 10.1007/s41783-024-00165-2.
17. Al-Hourani and D. Weraikat, “A Systematic Review of Artificial intelligence (AI) and Machine Learning (ML) in Pharmaceutical Supply Chain (PSC) Resilience: Current trends and future directions,” Sustainability, vol. 17, no. 14, p. 6591, Jul. 2025, doi: 10.3390/su17146591.