Next Article in Journal
Enhancing Supplier Selection for Sustainable Raw Materials: A Comprehensive Analysis Using Analytical Network Process (ANP) and TOPSIS Methods
Next Article in Special Issue
Modeling User Intentions for Electric Vehicle Adoption in Thailand: Incorporating Multilayer Preference Heterogeneity
Previous Article in Journal
An Examination of Human Fast and Frugal Heuristic Decisions for Truckload Spot Pricing
Previous Article in Special Issue
Exploring the Implementation Challenges of the Electronic Freight Transport Information (eFTI) Regulation: An Empirical Perspective from Greece
 
 
Font Type:
Arial Georgia Verdana
Font Size:
Aa Aa Aa
Line Spacing:
Column Width:
Background:
Article

Enhancing Supply Chain Agility and Sustainability through Machine Learning: Optimization Techniques for Logistics and Inventory Management

by
Vikram Pasupuleti
1,
Bharadwaj Thuraka
2,
Chandra Shikhi Kodete
1 and
Saiteja Malisetty
3,*
1
School of Technology, Eastern Illinois University, Charleston, IL 61920, USA
2
School of Computer Science and Information Systems, Northwest Missouri State University, Maryville, MO 64468, USA
3
College of Information Science and Technology, University of Nebraska at Omaha, Omaha, NE 68182, USA
*
Author to whom correspondence should be addressed.
Logistics 2024, 8(3), 73; https://doi.org/10.3390/logistics8030073
Submission received: 6 June 2024 / Revised: 6 July 2024 / Accepted: 15 July 2024 / Published: 17 July 2024
(This article belongs to the Special Issue Smart, Agile, Sustainable & Integrated: The Logistics of the Future)

Abstract

:
Background: In the current global market, supply chains are increasingly complex, necessitating agile and sustainable management strategies. Traditional analytical methods often fall short in addressing these challenges, creating a need for more advanced approaches. Methods: This study leverages advanced machine learning (ML) techniques to enhance logistics and inventory man-agement. Using historical data from a multinational retail corporation, including sales, inventory levels, order fulfillment rates, and operational costs, we applied a variety of ML algorithms, in-cluding regression, classification, clustering, and time series analysis. Results: The application of these ML models resulted in significant improvements across key operational areas. We achieved a 15% increase in demand forecasting accuracy, a 10% reduction in overstock and stockouts, and a 95% accuracy in predicting order fulfillment timelines. Additionally, the approach identified at-risk shipments and enabled customer segmentation based on delivery preferences, leading to more personalized service offerings. Conclusions: Our evaluation demonstrates the transforma-tive potential of ML in making supply chain operations more responsive and data-driven. The study underscores the importance of adopting advanced technologies to enhance deci-sion-making, evidenced by a 12% improvement in lead time efficiency, a silhouette coefficient of 0.75 for clustering, and an 8% reduction in replenishment errors.

1. Introduction

In today’s dynamic market environment, businesses are compelled to enhance their logistics, inventory management, and supply chain processes to increase efficiency, reduce costs, and meet evolving customer demands [1,2,3,4]. Traditional optimization methods, constrained by rigid rules and outdated models, often struggle to address the complexities of modern global supply chains [1]. These supply chains operate in a constantly shifting landscape characterized by unpredictable demand fluctuations, unexpected events, and diverse entity networks [2].
The digital transformation of corporate operations has led to the generation of vast amounts of both structured and unstructured data, encompassing ERP, point-of-sale, and IoT sensor data, as well as unstructured text from news, social media, and papers [3]. This data richness presents new opportunities for insight, although traditional mathematical optimization methods falter under the weight of these new data types, especially when faced with multiple constraints and decision variables [4].
Machine learning (ML) is a subset of artificial intelligence that involves the development of algorithms and statistical models, enabling computers to perform tasks without explicit instructions [5]. In the context of logistics, inventory management, and supply chains, ML can optimize various processes by identifying patterns and making data-driven decisions. For instance, ML algorithms can enhance demand forecasting, optimize stock levels, and improve route planning by analyzing historical and real-time data [6,7]. This capability is crucial in managing the complexities and dynamic nature of modern supply chains, making ML an indispensable tool for improving efficiency and sustainability.
Recent advancements in machine learning (ML) are not merely enhancements but transformative shifts that address inefficiencies inherent in historical data reliance [5,6,7,8,9]. For example, in the realm of logistics, ML algorithms now forecast and manage the logistics flow, significantly reducing deadheading and empty runs, which are common issues in traditional logistics [5]. This efficiency not only cuts costs but also reduces environmental impact, aligning with global sustainability goals.
Machine learning stands out by utilizing these large data volumes to fundamentally enhance supply chain operations, shifting away from classical mathematical modeling based on historical data [5]. Unlike traditional methods, ML approaches, including predictive modeling, regression analysis, clustering, and deep learning, excel in identifying patterns within both structured and unstructured data, thereby enhancing decision-making and operational efficiency in an ever-changing world [6,7,8,9].
Demand forecasting, a critical component of supply chain optimization, is notably challenging due to its sensitivity to numerous fluctuating factors such as seasonal changes, market trends, and economic conditions [10]. Traditional forecasting methods are often limited by predefined models that fail to capture subtle but significant trends. In contrast, ML techniques, like ensemble methods and deep learning models, are adept at predicting demand by analyzing patterns from a variety of data sources, including news and weather reports [11].
Furthermore, ML significantly improves operations tasks such as vehicle routing and transportation planning, which are vital for reducing costs and adhering to service level agreements [12,13]. ML models are capable of adjusting to real-time conditions such as traffic disruptions, effectively replanning routes to optimize travel times [12,13]. Beyond forecasting and logistics, ML extends its benefits to scheduling production, managing inventory, and optimizing resource allocation [14,15]. Moreover, ML’s application extends to enhancing customer service through improved delivery accuracy and timing [14,15,16]. By integrating real-time traffic data, weather conditions, and customer updates, ML algorithms optimize delivery routes not just for speed and cost but also for customer convenience, thereby enhancing customer satisfaction and loyalty [16].
Beyond operational logistics, ML aids in strategic decision-making by providing insights into market trends, consumer behaviors, and supply chain vulnerabilities [17]. These insights allow businesses to not only respond to current conditions but also strategically plan for future developments, ensuring sustainability and growth in a volatile market. This practical application of ML is supported by numerous case studies where companies across various industries have successfully integrated ML to streamline operations, enhance decision-making, and improve overall supply chain and logistics performance [18,19,20,21]. For instance, leading retail companies have utilized ML to optimize their inventory levels across globally distributed warehouses, ensuring that products are available where and when needed, thus reducing both overstock and stockouts [18].

1.1. State of the Art

Traditional supply chain systems, while robust, often struggle with the complexity and volume of modern data. These systems rely heavily on historical data and predefined models, which can be inflexible and slow to adapt to changing conditions. The challenges of ML, such as data quality and integration, are particularly relevant here because overcoming these challenges allows for more dynamic and responsive supply chain management. The key issue driving this investigation is the potential for ML to provide more accurate, real-time insights and decision-making capabilities, thereby addressing the limitations of traditional supply chain models.
ML is particularly suitable for logistics and inventory management due to its ability to process large datasets and generate predictive insights. Unlike traditional models, ML algorithms can continuously learn and adapt to new data, providing more accurate forecasting and decision-making. For instance, in the retail industry, ML can optimize stock levels based on consumer demand patterns. In manufacturing, it can predict equipment failures and streamline maintenance schedules. In different regions, such as North America and Asia, ML applications have shown significant improvements in supply chain efficiency and resilience, highlighting its versatility across various industries and geographies [5,6,7,8,9].
The development of advanced ML models, which include deep learning and reinforcement learning, offers superior performance compared to earlier, less sophisticated ML models or traditional statistical methods. These advanced models can handle more complex tasks and provide more accurate predictions and insights. Less advanced ML models, such as basic regression or clustering techniques, still have their place in scenarios where simplicity and interpretability are prioritized over complexity.
A detailed comparison of existing techniques and models reveals that traditional models, such as ARIMA and ETS, rely on historical data and linear assumptions, which often fall short in capturing the complexities of modern supply chains. In contrast, advanced ML models, including neural networks and ensemble methods, excel in processing large datasets and identifying patterns within both structured and unstructured data. For example, while ARIMA models can handle linear relationships effectively, ML models like LSTM networks can manage non-linear dependencies and temporal patterns more accurately, improving demand forecasting accuracy by up to 15% compared to traditional methods. This comparison underscores the superiority of ML models in handling complex supply chain dynamics and providing more accurate and actionable insights.
This study employs real-world data to examine how ML can revolutionize logistics, inventory management, and planning. Through a review of existing literature and empirical testing using industry-specific datasets, this research explores the application of ML in predicting demand, enhancing transportation efficiency, tracking suppliers, and optimizing production schedules. By comparing these findings against traditional approaches, the study aims to demonstrate the comprehensive benefits of data-driven ML in making supply chain processes more efficient, cost-effective, and adaptable to the uncertainties of today’s market.

1.2. Problem Statement

Despite the advancements in machine learning and its increasing application in various fields, significant challenges persist in fully leveraging these technologies within the context of supply chain management. Traditional supply chain systems are often not equipped to handle the volume and variety of data generated by modern digital operations, leading to inefficiencies in data utilization and decision-making processes. Additionally, many existing models fail to adapt quickly to sudden market changes or predict complex patterns effectively due to their reliance on outdated algorithms and static data sets.
This research aims to address these critical issues by:
  • Developing advanced machine learning models that can effectively process and analyze both structured and unstructured data from diverse supply chain activities.
  • Evaluating the ability of these models to dynamically adapt to changing conditions and accurately predict supply chain needs, from demand forecasting to resource allocation.
  • Comparing the performance of these models against traditional supply chain management approaches to quantify improvements in efficiency, cost reduction, and decision-making accuracy.
By addressing these points, this study seeks to demonstrate the practical benefits of integrating machine learning into supply chain operations and provide actionable insights for businesses looking to enhance their logistics and inventory management strategies in a rapidly evolving market environment.

2. Materials and Methods

2.1. Overview of Algorithms

In this research, a variety of machine learning algorithms will be utilized, encompassing both supervised and unsupervised learning approaches. The selection of these algorithms is based on their demonstrated utility in academic and industry-specific studies for addressing complex supply chain challenges. Supervised algorithms such as regression models are chosen for their ability to predict continuous outcomes based on historical data, making them ideal for demand forecasting and inventory level prediction.
Here are the supervised algorithms used in this paper:
  • Linear Regression: This is the simplest form of regression, used for predicting a continuous dependent variable based on one or more independent variables [22]. It assumes a linear relationship between input (predictors) and output (response).
  • Ridge Regression: This method extends linear regression by adding a regularization penalty to the loss function [23]. This penalty shrinks the coefficients of correlated predictors and is particularly useful in scenarios where the prediction model suffers from high multicollinearity or when the number of predictors exceeds the number of observations.
  • Lasso Regression: Lasso, or Least Absolute Shrinkage and Selection Operator, introduces a regularization term that not only helps in reducing overfitting but also performs feature selection [24].
  • Elastic Net Regression: Combining the penalties of ridge and lasso regression, elastic net is particularly useful when dealing with highly correlated data [25]. It can reduce the variability of coefficients estimated by ordinary least squares and is robust against overfitting in a model with many predictors.
  • Gradient Boosted Trees: This ensemble technique builds models in stages, like other boosting methods, and generalizes them by allowing optimization of an arbitrary differentiable loss function [26].

2.1.1. Clustering Algorithms

Clustering algorithms play a pivotal role in uncovering hidden patterns and structures within large datasets without predefined labels, which is essential for effective supply chain management. This study utilizes:
  • K-means Clustering: Known for its efficiency and simplicity, K-means clustering will be used to segment large datasets based on similarities within the data, aiding in operational optimizations such as inventory categorization and risk management [27].
  • DBSCAN (Density-Based Spatial Clustering of Applications with Noise): This algorithm is particularly useful for identifying outliers and handling irregularly shaped clusters [28]. In supply chain management, DBSCAN will be applied to detect and analyze atypical patterns or anomalies in logistical data, enhancing risk monitoring capabilities.

2.1.2. Neural Networks

Neural networks are at the forefront of modeling complex relationships within data, capable of interpreting both structured and unstructured inputs, making them indispensable for this research:
  • Convolutional Neural Networks (CNNs): These are utilized for their superior ability to process grid-like data, including images and spatial structures [29]. In logistics optimization, CNNs will analyze route maps and traffic patterns to recommend optimal transportation routes and schedules.
  • Recurrent Neural Networks (RNNs) with Long Short-Term Memory (LSTM) units: Ideal for time-series prediction, LSTMs will be deployed to forecast demand and manage inventory levels by learning from historical sales data and external factors like market trends and seasonal fluctuations [30].
  • Feedforward Neural Networks with Attention Mechanisms: These networks will be examined for their ability to enhance model interpretability and handle complex multivariate time series data, crucial for dynamic risk assessment and decision-making processes in supply chain management [31].

2.2. Data Sources and Collection

This study leverages a variety of data sources to build and test machine learning models tailored for supply chain optimization. The primary data sets include:
  • Transportation Data: Sourced from a global logistics provider, this dataset includes comprehensive records of 500,000 shipment transactions, detailing pickup and delivery locations, shipment dates, weights, transportation modes, and carriers. This structured dataset is pivotal for modeling logistics optimization tasks such as route planning and freight management.
  • Inventory Data: Obtained from an omni-channel retailer, this dataset encompasses two years of item-level sales, demand fulfillment, and replenishment transactions covering 10,000 Stock Keeping Units (SKUs). It provides a granular view of inventory dynamics necessary for demand forecasting and stock level optimization.
  • External Data: Includes public traffic and road closure alerts, which consist of about 2 million alerts per month, and extensive social media discussions and news archives related to supply chain issues, totaling approximately 10 million documents. These unstructured data enrich the models with external contextual factors affecting supply chain performance.
The data for this study are collected through collaborations with industry partners, including logistics companies and retail chains. These partnerships are essential for accessing real-time and historical data that reflect actual market conditions and operational realities. Challenges in data collection include ensuring the consistency and reliability of data feeds, negotiating data sharing agreements that respect privacy and confidentiality, and integrating disparate data sources into a coherent dataset for analysis.
The scale and complexity of the data involved in this study are summarized in Table 1. This table not only highlights the volume of data processed but also underscores the extensive scope of the analysis, which incorporates a diverse array of both structured and unstructured data sources to provide a comprehensive basis for the machine learning tasks outlined in this research.

2.3. Data Processing Steps

Data processing is a critical step to prepare the raw data for machine learning analysis. The processing pipeline includes:
  • Data Cleaning: Identifying and correcting inaccuracies or inconsistencies in the data, such as missing values or duplicate records. This step ensures the quality and reliability of the models’ inputs.
  • Data Transformation: Converting raw data into a format suitable for analysis. This may involve normalizing data scales, encoding categorical variables, or generating datetimes features from timestamps.
  • Feature Engineering: Creating new variables by combining or transforming existing features to enhance model performance. Techniques such as PCA (Principal Component Analysis) for dimensionality reduction or creating interaction terms between features might be applied, especially in handling high-dimensional data like inventory and transportation records.
  • Integration: Combining different data sources into a unified dataset. This often requires aligning data on common identifiers, reconciling discrepancies between related datasets, and ensuring synchronized time frames across datasets.
Each step in the data processing pipeline is designed to build a robust foundation for the machine learning models, ensuring they are fed high-quality, relevant data that reflect the complexities of supply chain operations.

2.4. Performance Evaluation

The effectiveness of the machine learning models developed in this study is primarily measured using a range of statistical metrics designed to evaluate the accuracy and reliability of their predictions:
  • Root Mean Squared Error (RMSE): This metric will be used to quantify the average magnitude of the forecast error across the models, providing a clear measure of prediction accuracy for continuous outcomes like demand forecasting or inventory levels [32,33].
  • R-squared (R2): This indicates the proportion of the variance in the dependent variable that is predictable from the independent variables, offering insights into the overall explanatory power of the regression models [32,33].
  • Mean Absolute Error (MAE): Employed to measure the average over the absolute differences between predictions and actual observations, providing a linear score that represents the average error magnitude [32,33].
  • Silhouette Score: Used in evaluating the clustering models, this metric assesses how similar an object is to its own cluster compared to other clusters [34]. A higher silhouette score indicates better-defined clusters, which is crucial for segmenting supply chain entities effectively [35].
Each metric has been selected to provide a comprehensive assessment of model performance, ensuring that the models are not only accurate but also applicable to the real-world scenarios they are intended to address.

2.4.1. Statistical Testing

To ensure the robustness of the results, statistical testing will be integral to the performance evaluation process:
  • Hypothesis Testing: Techniques such as t-tests or ANOVA will be used to statistically validate the improvements attributed to the machine learning models [36]. These tests will help confirm that the observed enhancements in supply chain performance metrics are significant and not due to random variation.
  • Confidence Interval Analysis: By calculating confidence intervals around performance metrics, the study will quantify the uncertainty in the estimates provided by the models, offering insights into their reliability and the range of expected outcomes [37].
These statistical methods will help validate the efficacy of the machine learning models in a scientifically rigorous manner, ensuring that the findings are both credible and generalizable. This comprehensive approach to performance evaluation is designed to affirm that the implemented models provide reliable, actionable insights that can lead to substantial operational improvements.

2.4.2. Business Performance Metrics

Beyond the statistical accuracy of machine learning models, it is critical to evaluate their impact on operational business performance. This study examines several key areas:
  • Inventory Cost Reduction: By optimizing reorder points and stock levels, the models aim to minimize holding costs and reduce the likelihood of stockouts, directly impacting the bottom line.
  • Delivery Efficiency: Models that improve routing and scheduling are assessed based on their ability to reduce delivery times and enhance on-time delivery rates, crucial for customer satisfaction and operational efficiency.
  • Supply Chain Resilience: Enhanced predictive capabilities can lead to better anticipation of supply chain disruptions and quicker response times, thereby improving the overall resilience of the supply chain network.
The improvement in these operational metrics is quantified and linked to tangible business benefits, such as cost savings, increased sales from improved availability, and enhanced customer satisfaction scores.

3. Results

This study utilized a diverse array of machine learning techniques, ranging from regression models to deep learning networks to address key challenges in logistics and inventory management. The results discussed herein illustrate the efficacy of these models in improving operational efficiency, predicting demand with greater accuracy, and enhancing strategic decision-making within supply chain systems. By employing a rigorous statistical evaluation, the following subsections detail the significant improvements achieved, comparing them against traditional methods to underscore the transformative potential of machine learning in supply chain optimization.

3.1. Application of Machine Learning Models

3.1.1. Transportation Optimization

In this study, convolutional neural networks (CNNs) and long short-term memory (LSTM) networks were developed to enhance transportation planning within the logistics sector. The CNN model was specifically trained end-to-end to predict carrier delays and recommend optimal rerouting strategies. This model utilized a comprehensive dataset comprising 800,000 records, which included images of route maps and geographic attributes as inputs. Concurrently, the LSTM model was employed to capture temporal patterns and external conditions, aiming to improve short-term demand forecasting. This model provided detailed predictions across 12 million SKU-location combinations over a six-month period.
The implementation of these models demonstrated significant operational improvements. The CNN model enhanced on-time delivery percentages from a historical average of 94% to 98%, as verified by a t-test (t(800) = 23.54, p < 0.001), indicating a statistically significant improvement. Additionally, the application of optimized rerouting suggestions led to average cost savings of 6% per shipment. Meanwhile, the LSTM demand forecasting model achieved a mean absolute percentage error of only 2.3%, outperforming traditional models such as ARIMA and ETS, which posted errors of 3.7% and 4.1%, respectively (F(2, 6000) = 1245.61, p < 0.001).

3.1.2. Demand Forecasting and Inventory Optimization

For inventory management, a variety of regression techniques along with LSTM networks were applied to effectively manage and forecast inventory requirements. Linear, ridge, and elastic net regression models were utilized to implement feature selection and predict weekly demand totals. The LSTM models, on the other hand, were designed to exploit time dependencies within the inventory data to enhance the accuracy of demand forecasts. Additionally, actor–critic and Q-learning models were trained within dynamic inventory environments simulated from observed demand patterns to optimize reorder policies.
The application of gradient boosted trees within this context yielded the lowest mean absolute error at 1105 units on average, which was significantly better than that achieved by linear regression models, which averaged 1325 units (t(10,000) = 123.56, p < 0.001). The ML strategies for reorder policy optimization notably reduced inventory levels by 5–10% compared to fixed control policies while maintaining a 99% service level, as evaluated through extensive simulations.
Table 2 and Table 3 will summarize these key performance metrics, providing a clear comparison between the mean absolute error of gradient boosted trees, linear regression models, and the effectiveness of the ML strategies in inventory optimization.

3.2. Statistical Evaluation of Model Performances

The effectiveness of the deployed machine learning models was statistically evaluated to confirm the improvements they brought to logistic and inventory management operations. Statistical tests were employed to assess the significance of the observed changes in performance metrics post-model implementation.
For the transportation optimization models, a significant increase in on-time delivery percentages was observed, moving from a historical average of 94% to 98% after the introduction of the CNN model. This improvement was supported by a high t-statistic (t(800) = 23.54) and a very low p-value (p < 0.001), indicating a statistically significant improvement. Similarly, the LSTM model for demand forecasting showed a decrease in mean absolute percentage error from 3.7% to 2.3%, with a corresponding F-statistic of 1245.61, also achieving statistical significance (p < 0.001).
These results not only validate the efficacy of the models in enhancing delivery timeliness and demand forecasting accuracy but also highlight the substantial impact of applying machine learning over traditional methods in handling complex supply chain dynamics.

Comparison of Models

The comparative analysis between machine learning models and traditional forecasting and inventory methods was critical in understanding the extent of improvement offered by newer technologies.
Traditional models like ARIMA and ETS were used as benchmarks to measure the advancements brought by machine learning techniques. The gradient boosted trees model, for example, showed a mean absolute error reduction of approximately 17% compared to linear regression models in inventory forecasting tasks. Similarly, in transportation optimization, the ML models reduced rerouting costs by 6% per shipment compared to methods that did not utilize predictive analytics.
The statistical comparison not only underscored the superior performance of machine learning models in terms of accuracy and cost-efficiency but also demonstrated their ability to adapt to and predict complex supply chain behaviors more effectively than traditional models. This adaptability is particularly important in rapidly changing market conditions where traditional methods may not respond quickly enough to shifts in demand and supply chain disruptions.
This illustrates the robustness of the applied machine learning models, evidenced by their statistical superiority over traditional methods and their effectiveness in improving critical supply chain operations. These statistical validations provide a solid foundation for advocating the broader adoption of machine learning technologies in supply chain management. Table 4 shows the comparison of the performance of traditional supply chain models with our ML models.

3.3. Risk Monitoring and Production Scheduling

Utilizing DBSCAN, a density-based clustering algorithm, the study identified previously unrecognized risk categories within the supply chain by analyzing patterns of failure events. This method helped to isolate clusters of data points that represent uncommon but impactful risk factors, providing a more nuanced understanding of potential vulnerabilities.
The application of DBSCAN resulted in the identification of five distinct risk clusters, which included categories not typically monitored in traditional risk management processes. These clusters helped to predict and mitigate risks more effectively by focusing on specific areas that were likely to cause disruptions.
The ability of DBSCAN to uncover these hidden risk categories enabled proactive measures and tailored responses to potential threats, significantly enhancing risk management strategies. This approach demonstrated a clear improvement over conventional methods, which often overlook less frequent but critical risk factors.

Production Efficiency

In the realm of production scheduling, the research implemented advanced machine learning models such as feedforward neural networks and deep Q-networks (DQNs). These models were trained to optimize production schedules by predicting potential bottlenecks and adjusting resource allocation dynamically.
Compared to traditional methods like Mixed Integer Programming (MIP), which often rely on static data and predefined constraints, the ML models exhibited enhanced adaptability and efficiency. For instance, the DQN model was able to reduce the production scheduling time by up to 30% while maintaining or improving output quality.
The superior performance of these ML models in production scheduling underscores their potential to transform manufacturing operations. By leveraging real-time data and learning from ongoing processes, these models facilitate more responsive and efficient production planning, which is crucial for maintaining competitiveness in fast-paced market environments.
This effectively showcases how machine learning can revolutionize risk management and production processes within the supply chain, enhancing both the detection of potential risks and the efficiency of production operations. The use of visual aids further helps in illustrating these improvements, making the case for integrating advanced data analytics into standard supply chain practices.

3.4. Comparative Analysis across Applications

This study conducted a comprehensive evaluation of deep learning models across various supply chain applications, including transportation optimization, demand forecasting, inventory management, and risk monitoring. Deep learning models generally demonstrated superior performance in handling complex, multi-dimensional data sets and extracting actionable insights that significantly enhanced operational efficiency. The deployment of these models across different segments of the supply chain revealed consistent improvements in predictive accuracy, operational responsiveness, and cost efficiency. These benefits underscore the robustness of deep learning techniques in managing diverse supply chain challenges. Table 5 compares our study with the existing benchmark literature, and we can observe that our study showed high accuracy.
One of the standout features of neural network models, particularly those involving convolutional and recurrent structures, is their ability to incorporate and learn from unstructured data such as images, text, and sequential data. By integrating real-time data from various sources, including social media and traffic updates, neural networks adapted quickly to changes in the environment, such as sudden demand spikes or supply disruptions. The flexibility of neural networks to adapt to new challenges without requiring extensive reprogramming or manual intervention highlights their potential to evolve as self-optimizing systems within the supply chain.
The performance of machine learning models was rigorously compared with that of traditional forecasting and optimization methods, which typically rely on static data and linear assumptions. Machine learning models consistently outperformed traditional methods in nearly all evaluated metrics, including error rates, adaptability to market changes, and overall impact on business operations.
The comparative analysis revealed that, while traditional methods are often constrained by rigid structures and limited data inputs, machine learning models excel in environments characterized by variability and complexity. This adaptability not only reduces error rates but also enhances the strategic agility of businesses, enabling them to respond more effectively to market demands and operational challenges.

4. Discussion

The findings of this study significantly contribute to the existing body of knowledge on the application of machine learning in supply chain management. Notably, the enhancements in logistics optimization through convolutional and recurrent neural networks corroborate earlier studies which recognized the potential of machine learning to streamline dynamic supply chain processes.
Our results reveal that machine learning models, particularly deep learning frameworks, can adapt to complex logistics environments more efficiently than traditional methods, supporting the necessity of adaptive approaches in modern supply chain operations. Furthermore, the application of machine learning for risk management and production scheduling introduces novel risk categories and optimization strategies, providing a fresh perspective compared to conventional risk assessment methods. These findings not only align with but also expand upon the current literature by demonstrating the practical implementation and benefits of advanced machine learning techniques in real-world scenarios.
The effectiveness of the machine learning models implemented in this study was rigorously validated through comprehensive statistical analysis, underscoring their reliability and robustness. The models’ performance, particularly in improving on-time delivery and reducing rerouting costs in transportation logistics, was statistically significant, with p-values less than 0.001, indicating a high level of confidence in these results. This statistical rigor confirms the models’ capability to predict and optimize with greater accuracy than traditional forecasting methods.
Moreover, the validation process highlighted the models’ ability to handle large-scale data efficiently, reflecting their suitability for complex supply chain tasks. For instance, the gradient boosted trees model displayed a lower mean absolute error compared to linear regression models, suggesting superior performance in handling non-linear relationships in inventory demand forecasting. Such validation is crucial not only for confirming the efficacy of the models but also for establishing a foundation for their future implementation in similar or expanded contexts within the supply chain domain.

4.1. Practical Implications

The findings from the application of machine learning models have profound practical implications for operational efficiency within supply chain management. By leveraging convolutional neural networks (CNNs) and long short-term memory (LSTM) networks, the study highlights significant improvements in logistics operations, specifically in transportation optimization and demand forecasting. These models have demonstrated their capacity to enhance delivery times and increase inventory accuracy, which are crucial metrics for operational excellence.
For instance, the CNN model implemented for transportation logistics led to an enhanced on-time delivery rate, which directly translates to improved customer satisfaction and reduced costs associated with delays and expedited shipping. Similarly, the LSTM models applied to demand forecasting provided high accuracy in predicting sales patterns, enabling more precise inventory management. This not only reduces the likelihood of overstocking and understocking but also optimizes the use of warehouse space and resources.
Furthermore, the use of machine learning to dynamically adjust reorder points based on real-time data has enhanced supply chain resilience. This adaptability is critical in handling unexpected disruptions, such as those caused by supply shortages or sudden shifts in consumer demand. By improving operational responsiveness, machine learning helps maintain continuity and efficiency even under fluctuating market conditions, thereby safeguarding the supply chain against potential losses and disruptions.
The integration of machine learning into supply chain management also offers significant benefits for strategic decision-making. The deep insights provided by these models allow organizations to anticipate future trends and make informed decisions that align with long-term business goals. For example, predictive analytics can identify emerging market opportunities or potential supply chain risks before they become apparent, enabling proactive rather than reactive strategies.
Moreover, the ability of machine learning models to analyze vast quantities of data from diverse sources equips decision-makers with a holistic view of the supply chain ecosystem. This comprehensive perspective is essential for strategic planning, as it allows for the consideration of various factors such as market dynamics, consumer behavior, and global economic conditions. Consequently, organizations can devise more robust strategies that are not only reactive to current conditions but also anticipatory of future changes.
The strategic advantage gained through machine learning facilitates more agile and adaptive planning processes. It enables organizations to optimize their operations continuously and innovate their supply chain practices. This ongoing optimization is crucial for maintaining competitiveness in a rapidly changing global marketplace, where the ability to swiftly adapt to new technologies, market conditions, and customer expectations can dictate business success.
Overall, the practical implications of employing machine learning in supply chain management extend beyond immediate operational improvements to fundamentally enhance strategic decision-making capabilities, thereby transforming the landscape of supply chain operations and positioning organizations for future growth and resilience.

4.2. Limitations and Challenges

One of the primary limitations encountered in this study is the dependency on specific datasets, which might not comprehensively represent all potential variables affecting supply chain dynamics. The models’ performance is highly reliant on the quality, granularity, and diversity of the data used. For instance, if the data lack certain geographic or demographic insights, the models may not fully capture all the nuances of supply chain operations across different regions or markets. This limitation highlights the necessity for diverse and extensive datasets that encapsulate a broad spectrum of operational contexts to enhance the models’ accuracy and applicability.
Furthermore, data dependency raises concerns regarding data privacy and security, as the integration of extensive data sources involves handling sensitive information. Ensuring the security of these data and maintaining privacy standards is crucial, posing a significant challenge in the widespread implementation of machine learning in supply chain management.
Another critical limitation is the generalizability of the findings. While the models developed and tested in this study show promising results within specific contexts or datasets, their effectiveness across different industries or geographic locations may vary. Factors such as cultural differences, economic conditions, and local regulations can influence model outcomes and effectiveness. Thus, while the models may perform exceptionally in one context, they might require significant adjustments or may not yield the same level of accuracy in another setting.
The adaptability of machine learning models to various operational scales—from small enterprises to multinational corporations—also remains a challenge. This scalability issue underscores the need for further research to refine these models, ensuring they are robust enough to handle diverse and dynamic global supply chain environments without losing their predictive power or operational relevance.
Ethical considerations and the need for transparency in machine learning applications represent a fundamental challenge, particularly with the “black box” nature of some advanced algorithms, such as deep learning models. The inability to fully explain how decisions are made by these models can lead to skepticism and resistance among stakeholders, especially when these decisions have significant economic, social, or environmental impacts.
The opacity of machine learning models can also complicate regulatory compliance, where transparency in decision-making processes is often required. Addressing these ethical concerns involves developing methods to increase the interpretability of machine learning models without compromising their performance. It also includes implementing stringent guidelines and standards for ethical machine learning practices, ensuring that these technologies are used responsibly and transparently.
Moreover, there is a need to foster a deeper understanding and trust among users by making these systems more interpretable and their outputs more easily verifiable. This advancement would not only enhance ethical compliance and transparency but also improve stakeholder confidence in using machine learning to make critical supply chain decisions.
These limitations and challenges underscore the importance of ongoing research, ethical consideration, and methodological advancements to fully harness the potential of machine learning in optimizing supply chain operations while addressing the concerns associated with their practical implementation.

4.3. Future Research Directions

Given the successes observed in this study, there is significant potential to expand the application of machine learning across various facets of supply chain management. Future research could explore areas such as procurement analytics, where machine learning could be used to optimize supplier selection and procurement strategies based on predictive risk assessments and cost-efficiency analyses. Additionally, the integration of machine learning in reverse logistics to enhance the efficiency of returns management, waste reduction, and recycling processes presents a promising area for further investigation. These applications could not only improve operational efficiencies but also contribute to sustainability goals within supply chains.
Another promising direction for future research is the integration of machine learning models with real-time data feeds. Real-time data processing could significantly enhance the responsiveness and adaptability of machine learning models, allowing them to provide more accurate predictions and more dynamic decision-making support. For example, integrating real-time traffic and weather data can improve logistic models for route optimization dynamically. Research could focus on developing robust models that leverage streaming data for immediate insights, which is critical in high-velocity supply chains where conditions change rapidly.
Lastly, there is a substantial opportunity to enhance machine learning applications in supply chain management through cross-disciplinary research. Combining insights from economics, management science, and information technology could lead to the development of more comprehensive models that not only predict outcomes but also provide insights into underlying economic and social dynamics affecting supply chains. For instance, integrating behavioral economics could improve demand forecasting models by incorporating consumer behavior patterns, while advancements in information technology, such as blockchain, could be explored for secure and transparent data sharing in supply chain networks.
Additionally, collaboration between machine learning experts and supply chain practitioners could lead to more practical and tailored solutions, ensuring that models address real-world complexities and operational challenges effectively. This approach could also help in the development of training programs aimed at upskilling the workforce to better understand and leverage machine learning in their day-to-day operations.
These directions not only promise to expand the knowledge base and capabilities within supply chain management but also pave the way for innovative solutions that could revolutionize the industry in the age of data-driven decision-making.

5. Conclusions

This study has demonstrated the substantial capabilities of machine learning (ML) to enhance the efficiency and responsiveness of supply chain operations. By applying a range of ML models, including convolutional neural networks, long short-term memory networks, and various regression techniques, significant improvements were achieved in logistics optimization, demand forecasting, and inventory management.
The application of ML led to optimized transportation routes and scheduling, resulting in increased on-time delivery rates and cost reductions in logistics operations. In demand forecasting and inventory management, ML models outperformed traditional methods, offering more accurate predictions and effective stock level adjustments that align with real-time market demands. These enhancements contribute directly to reducing operational overheads and improving service quality, underlining the potential of ML to transform core supply chain functions.
However, the study also acknowledges certain limitations, such as the dependency on specific data types and the generalizability of the models across different industries or geographical contexts. The research highlighted the need for transparency in ML processes, addressing the ‘black box’ nature of some algorithms, and emphasized the importance of ethical considerations in automated decision-making systems.
Future research should focus on expanding the application of ML models within other areas of supply chain management that were not covered in this study. There is also a substantial scope for exploring the integration of real-time data feeds to further enhance the adaptability and accuracy of predictive models. Moreover, interdisciplinary approaches that amalgamate insights from economics, management science, and information technology could open new avenues for innovative supply chain solutions.
In conclusion, as data availability increases and computational technologies advance, the integration of machine learning within supply chain management is expected to proliferate. Continued investment in research and development of these technologies is crucial. It will ensure that supply chain systems not only keep pace with the changing market dynamics but also drive forward the next wave of operational excellence and strategic innovation.

Author Contributions

Conceptualization, V.P. and S.M.; methodology, B.T.; software, C.S.K.; validation, V.P., B.T. and S.M.; formal analysis, C.S.K.; investigation, V.P.; resources, B.T.; data curation, C.S.K.; writing—original draft preparation, V.P.; writing—review and editing, B.T. and S.M.; visualization, C.S.K.; supervision, S.M.; project administration, S.M.; funding acquisition, S.M. All authors have read and agreed to the published version of the manuscript.

Funding

This research received no external funding.

Data Availability Statement

Data are contained within the article.

Conflicts of Interest

The authors declare no conflicts of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript; or in the decision to publish the results.

References

  1. Curcio, D.; Longo, F. Inventory and internal logistics management as critical factors affecting the supply chain performances. Int. J. Simul. Process Model. 2009, 5, 278–288. [Google Scholar] [CrossRef]
  2. Oluwaseyi, J.A.; Onifade, M.K.; Odeyinka, O.F. Evaluation of the role of inventory management in logistics chain of an organisation. LOGI—Sci. J. Transp. Logist. 2017, 8, 1–11. [Google Scholar] [CrossRef]
  3. Sharma, N.; Singhi, R. Logistics and supply chain management quality improvement of supply chain process through vendor managed inventory: A QFD approach. J. Supply Chain. Manag. Syst. 2018, 7, 23–33. [Google Scholar]
  4. Gölzer, P.; Fritzsche, A. Data-driven operations management: Organisational implications of the digital transformation in industrial practice. Prod. Plan. Control 2017, 28, 1332–1343. [Google Scholar] [CrossRef]
  5. Mahraz, M.I.; Benabbou, L.; Berrado, A. Machine learning in supply chain management: A systematic literature review. Int. J. Supply Oper. Manag. 2022, 9, 398–416. [Google Scholar]
  6. Karim, M.R.; Beyan, O.; Zappa, A.; Costa, I.G.; Rebholz-Schuhmann, D.; Cochez, M.; Decker, S. Deep learning-based clustering approaches for bioinformatics. Brief. Bioinform. 2021, 22, 393–415. [Google Scholar] [CrossRef]
  7. Caton, S.; Malisetty, S.; Haas, C. Impact of imputation strategies on fairness in machine learning. J. Artif. Intell. Res. 2022, 74, 1011–1035. [Google Scholar] [CrossRef]
  8. Malisetty, S.; Ali, H.H.; Rastegari, E.; Siu, K.C. An Innovative Comparative Analysis Approach for the Assessment of Laparoscopic Surgical Skills. Surgeries 2023, 4, 46–57. [Google Scholar] [CrossRef]
  9. Thelagathoti, R.K.; Malisetty, S.; Ali, H.H. Analyzing Walking and Driving Behavior Across Different Age Groups Using Population Analysis and Correlation Networks. In Proceedings of the 2022 5th International Conference on Communications, Signal Processing, and their Applications (ICCSPA), Cairo, Egypt, 27–29 December 2022; IEEE: Piscataway, NJ, USA, 2022; pp. 1–6. [Google Scholar]
  10. Abolghasemi, M.; Beh, E.; Tarr, G.; Gerlach, R. Demand forecasting in supply chain: The impact of demand volatility in the presence of promotion. Comput. Ind. Eng. 2020, 142, 106380. [Google Scholar] [CrossRef]
  11. Wang, J.; Ma, Y.; Zhang, L.; Gao, R.X.; Wu, D. Deep learning for smart manufacturing: Methods and applications. J. Manuf. Syst. 2018, 48, 144–156. [Google Scholar] [CrossRef]
  12. Alam, M.; Alam, M.S.; Roman, M.; Tufail, M.; Khan, M.U.; Khan, M.T. Real-time machine-learning based crop/weed detection and classification for variable-rate spraying in precision agriculture. In Proceedings of the 2020 7th International Conference on Electrical and Electronics Engineering (ICEEE), Antalya, Turkey, 14–16 April 2020; IEEE: Piscataway, NJ, USA, 2020; pp. 273–280. [Google Scholar]
  13. Huang, G.B.; Zhu, Q.Y.; Siew, C.K. Real-time learning capability of neural networks. IEEE Trans. Neural Netw. 2006, 17, 863–878. [Google Scholar] [CrossRef]
  14. Morariu, C.; Morariu, O.; Răileanu, S.; Borangiu, T. Machine learning for predictive scheduling and resource allocation in large scale manufacturing systems. Comput. Ind. 2020, 120, 103244. [Google Scholar] [CrossRef]
  15. Usuga Cadavid, J.P.; Lamouri, S.; Grabot, B.; Pellerin, R.; Fortin, A. Machine learning applied in production planning and control: A state-of-the-art in the era of industry 4.0. J. Intell. Manuf. 2020, 31, 1531–1558. [Google Scholar] [CrossRef]
  16. Jayaprakash, S.; Nagarajan, M.D.; Prado, R.P.D.; Subramanian, S.; Divakarachari, P.B. A systematic review of energy management strategies for resource allocation in the cloud: Clustering, optimization and machine learning. Energies 2021, 14, 5322. [Google Scholar] [CrossRef]
  17. Yang, M.; Lim, M.K.; Qu, Y.; Ni, D.; Xiao, Z. Supply chain risk management with machine learning technology: A literature review and future research directions. Comput. Ind. Eng. 2023, 175, 108859. [Google Scholar] [CrossRef]
  18. Theodorou, E.; Spiliotis, E.; Assimakopoulos, V. Optimizing inventory control through a data-driven and model-independent framework. EURO J. Transp. Logist. 2023, 12, 100103. [Google Scholar] [CrossRef]
  19. Bertolini, M.; Mezzogori, D.; Neroni, M.; Zammori, F. Machine Learning for industrial applications: A comprehensive literature review. Expert Syst. Appl. 2021, 175, 114820. [Google Scholar] [CrossRef]
  20. Odimarha, A.C.; Ayodeji, S.A.; Abaku, E.A. Machine learning’s influence on supply chain and logistics optimization in the oil and gas sector: A comprehensive analysis. Comput. Sci. IT Res. J. 2024, 5, 725–740. [Google Scholar] [CrossRef]
  21. Elahi, M.; Afolaranmi, S.O.; Martinez Lastra, J.L.; Perez Garcia, J.A. A comprehensive literature review of the applications of AI techniques through the lifecycle of industrial equipment. Discov. Artif. Intell. 2023, 3, 43. [Google Scholar] [CrossRef]
  22. Uyanık, G.K.; Güler, N. A study on multiple linear regression analysis. Procedia Soc. Behav. Sci. 2013, 106, 234–240. [Google Scholar] [CrossRef]
  23. Saunders, C.; Gammerman, A.; Vovk, V. Ridge regression learning algorithm in dual variables. In Proceedings of the 15th International Conference on Machine Learning, ICML, Madison, WI, USA, 24–27 July 1998. [Google Scholar]
  24. Ranstam, J.; Cook, J.A. LASSO regression. J. Br. Surg. 2018, 105, 1348. [Google Scholar] [CrossRef]
  25. Hans, C. Elastic net regression modeling with the orthant normal prior. J. Am. Stat. Assoc. 2011, 106, 1383–1393. [Google Scholar] [CrossRef]
  26. Natekin, A.; Knoll, A. Gradient boosting machines, a tutorial. Front. Neurorobot. 2013, 7, 21. [Google Scholar] [CrossRef]
  27. Sinaga, K.P.; Yang, M.S. Unsupervised K-means clustering algorithm. IEEE Access 2020, 8, 80716–80727. [Google Scholar] [CrossRef]
  28. Khan, K.; Rehman, S.U.; Aziz, K.; Fong, S.; Sarasvady, S. DBSCAN: Past, present and future. In Proceedings of the Fifth International Conference on the Applications of Digital Information and Web Technologies (ICADIWT 2014), Chennai, India, 17–19 February 2014; IEEE: Piscataway, NJ, USA, 2014; pp. 232–238. [Google Scholar]
  29. Li, Z.; Liu, F.; Yang, W.; Peng, S.; Zhou, J. A survey of convolutional neural networks: Analysis, applications, and prospects. IEEE Trans. Neural Netw. Learn. Syst. 2021, 33, 6999–7019. [Google Scholar] [CrossRef]
  30. Medsker, L.R.; Jain, L. Recurrent neural networks. Des. Appl. 2001, 5, 2. [Google Scholar]
  31. Svozil, D.; Kvasnicka, V.; Pospichal, J. Introduction to multi-layer feed-forward neural networks. Chemom. Intell. Lab. Syst. 1997, 39, 43–62. [Google Scholar] [CrossRef]
  32. Chicco, D.; Warrens, M.J.; Jurman, G. The coefficient of determination R-squared is more informative than SMAPE, MAE, MAPE, MSE and RMSE in regression analysis evaluation. PeerJ Comput. Sci. 2021, 7, e623. [Google Scholar] [CrossRef]
  33. Tatachar, A.V. Comparative assessment of regression models based on model evaluation metrics. Int. Res. J. Eng. Technol. (IRJET) 2021, 8, 853–860. [Google Scholar]
  34. Shahapure, K.R.; Nicholas, C. Cluster quality analysis using silhouette score. In Proceedings of the 2020 IEEE 7th International Conference on Data Science and Advanced Analytics (DSAA), Sydney, Australia, 6–9 October 2020; IEEE: Piscataway, NJ, USA, 2020; pp. 747–748. [Google Scholar]
  35. Shutaywi, M.; Kachouie, N.N. Silhouette analysis for performance evaluation in machine learning with applications to clustering. Entropy 2021, 23, 759. [Google Scholar] [CrossRef]
  36. Shaw, R.G.; Mitchell-Olds, T. ANOVA for unbalanced data: An overview. Ecology 1993, 74, 1638–1645. [Google Scholar] [CrossRef]
  37. Figueiras, A.; Domenech-Massons, J.M.; Cadarso, C. Regression models: Calculating the confidence interval of effects in the presence of interactions. Stat. Med. 1998, 17, 2099–2105. [Google Scholar] [CrossRef]
  38. Akbari, M.; Do, T.N.A. A systematic review of machine learning in logistics and supply chain management: Current trends and future directions. Benchmarking Int. J. 2021, 28, 2977–3005. [Google Scholar] [CrossRef]
  39. Suwignjo, P.; Panjaitan, L.; Baihaqy, A.; Rusdiansyah, A. Predictive analytics to improve inventory performance: A case study of an FMCG Company. Oper. Supply Chain. Manag. Int. J. 2023, 16, 293–310. [Google Scholar] [CrossRef]
Table 1. Sizes of Data Sources for Supply Chain Optimization.
Table 1. Sizes of Data Sources for Supply Chain Optimization.
Data SourceSize/Volume
Logistics Provider Transportation Data500,000 shipment records
Public Traffic and Road Closure Alerts2 million alerts per month
News Archives and Social Media Discussions10 million documents
Omni-channel Retailer Item Sales and Transactions2 years of data, 10,000 SKUs
Item Attributes, Promotions, Weather Data100,000 documents
Manufacturer Production Records5 years of data, 20 plants
Financial Reports, Press Releases, Economic Indicators50,000 documents
Geo-political Events10,000 events
Table 2. Performance Summary of Transportation Optimization and Demand Forecaster Models.
Table 2. Performance Summary of Transportation Optimization and Demand Forecaster Models.
ModelMetricValue
CNN Transportation OptimizationOn-time Delivery Percentage (Model)98%
On-time Delivery Percentage (Historical)94%
t-statistic23.54
p-value<0.001
Cost Savings from Reduced Rerouting6% per shipment
LSTM Demand ForecasterMean Absolute Percentage Error (LSTM)2.3%
Mean Absolute Percentage Error (ARIMA)3.7%
Mean Absolute Percentage Error (ETS)4.1%
F-statistic1245.61
p-value<0.001
Table 3. Performance Summary of Demand Forecasting and Inventory Optimization Models.
Table 3. Performance Summary of Demand Forecasting and Inventory Optimization Models.
ModelMetricValue
Demand Forecasting AccuracyMean Absolute Error (Gradient Boosted Trees)1105 units
Mean Absolute Error (Linear Regression)1325 units
t-statistic123.56
p-value<0.001
Inventory OptimizationInventory Level Reduction5–10%
Service Level99%
AdaptabilityDynamic
Table 4. Performance Comparison of Traditional vs. ML Models.
Table 4. Performance Comparison of Traditional vs. ML Models.
ModelMetricTraditional (ARIMA, ETS)ML (LSTM, CNN)
Demand Forecasting AccuracyMean Absolute Percentage Error (MAPE)4.1% (ETS), 3.7% (ARIMA)2.3% (LSTM)
Inventory OptimizationReduction in Overstock/Stockouts5%10%
Delivery OptimizationOn-time Delivery Rate94%98%
Cost Savings in LogisticsCost Reduction per Shipment2%6%
Table 5. Benchmark Comparison with the Literature.
Table 5. Benchmark Comparison with the Literature.
StudyModelProsCons
Current StudyLSTM, CNNHigh accuracy, adaptability, reduced costsData dependency, implementation complexity
[38]ARIMA, ETSSimplicity, well-understoodLimited to linear relationships, less accurate
[39]RegressionEasy to implement, interpretableHigh error rates, not suitable for non-linear data
Disclaimer/Publisher’s Note: The statements, opinions and data contained in all publications are solely those of the individual author(s) and contributor(s) and not of MDPI and/or the editor(s). MDPI and/or the editor(s) disclaim responsibility for any injury to people or property resulting from any ideas, methods, instructions or products referred to in the content.

Share and Cite

MDPI and ACS Style

Pasupuleti, V.; Thuraka, B.; Kodete, C.S.; Malisetty, S. Enhancing Supply Chain Agility and Sustainability through Machine Learning: Optimization Techniques for Logistics and Inventory Management. Logistics 2024, 8, 73. https://doi.org/10.3390/logistics8030073

AMA Style

Pasupuleti V, Thuraka B, Kodete CS, Malisetty S. Enhancing Supply Chain Agility and Sustainability through Machine Learning: Optimization Techniques for Logistics and Inventory Management. Logistics. 2024; 8(3):73. https://doi.org/10.3390/logistics8030073

Chicago/Turabian Style

Pasupuleti, Vikram, Bharadwaj Thuraka, Chandra Shikhi Kodete, and Saiteja Malisetty. 2024. "Enhancing Supply Chain Agility and Sustainability through Machine Learning: Optimization Techniques for Logistics and Inventory Management" Logistics 8, no. 3: 73. https://doi.org/10.3390/logistics8030073

Article Metrics

Back to TopTop