A systematic review of machine learning in logistics and supply chain management: current trends and future directions

Current processes often depend on unreliable sources of data and outdated IT systems, with coordination limited across functions. To be fair, certain CPG companies consistently follow data-based planning methods, but even they typically optimize decision making at the local level, rather than globally, and with limited ability to address potential disruptions in real time. In this course, we’ll learn about more advanced machine learning methods that are used to tackle problems in the supply chain. We’ll start with an overview of the different ML paradigms (regression/classification) and where the latest models fit into these breakdowns.

ML can be used to generate realistic what-if scenarios based on historical data, enabling businesses to evaluate the potential impact of various decisions and plans on supply chain performance. During the COVID-19 pandemic, one of the largest branded consumer food and beverage product companies in Asia sought to improve its supply chain performance through autonomous planning. The company had historically used traditional processes, including an annual budget plan for forecasting, and it made highly manual, rule-of-thumb decisions in areas such as inventory levels and dispatch planning. Response times were slow—the company typically required more than five days to create a demand plan, and more than two days to create a dispatch plan. The company wanted to use analytics more effectively so that it could react faster to changes in supply or demand and do so in the most profitable way.

Key Components of Supply Chain Network Optimization Driven by Machine Learning and Novel Technologies

Manufacturers should invest in tools that centralize order management, consolidate orders from multiple channels, and track product returns. Unreliable suppliers—those that deliver shipments late, short orders, or send the wrong materials—make it difficult for manufacturers to maintain the proper balance of inventory, neither too much nor too little. Some companies impose penalties on suppliers that fail to meet their commitments—for example, shortening grace periods for a late delivery before fees are imposed. But companies also need to reward suppliers that deliver reliably—for example, by paying bonuses.

Guo et al. (2020) developed a system for meat freshness monitoring using cross-reactive colorimetric barcode combinatorics and a DL algorithm. This system provides both scent fingerprint and fingerprint recognition via a smartphone APP interface. In another study, Cavallo et al. (2018) proposed a methodology based on DL for Non-destructive evaluation of packaged vegetable freshness using product images. Moreover, the other top keyword is the internet of things technology that provide companies with a huge volume of data and can be used, managed, and optimized effectively by DL techniques (Khan et al. 2020). Image processing and text mining are two main processes that can be efficiently done using DL algorithms. The keyword “Text mining” has also been used frequently in the collected material according to the three-field plot.

Logistics Management

Machine learning will combine this data to predict demand for specific goods and help to manage the sourcing and manufacture of those products. Machine learning can analyze this information and use the findings to enhance supply chain management (SCM). As a trusted partnership emerges over time, one of the key areas to be renegotiated is accounts payable terms. For example, vendors can extend payment due dates, a big benefit to manufacturers that may be waiting for inventory to sell. Early-payment discounts can also be negotiated, a win for suppliers and manufacturers alike.

Demand forecasting Demand forecasting is defined as the forecast of demands for the products in stock that comes from the customers wanting to buy these goods for their use (Thomopoulos 2015). Demand uncertainty encounters supply chains with many problems such as the bullwhip effect (Bousqaoui et al. 2021). These methods also help supply chains to optimize inventories, reduce costs, increase profit, and gain more customer loyalty (Kilimci et al. 2019). PepsiCo, a global food and beverage company, has utilized machine learning to optimize its production and inventory planning processes.

In the execution stage, managers align order management, warehouse and inventory management, and transportation logistics to ensure that products get to retailers or customers as quickly and reliably as possible at the lowest cost. For example, some manufacturers use transportation management software to improve visibility into product shipments, shortening shipping times and increasing customer satisfaction. Others use AI to detect anomalies caused by human error or machine failure and refine their processes based on that feedback. Having higher levels of abstraction, DL algorithms, as an advancement of artificial neural networks, are expected to improve the results (Mocanu et al. 2016) in comparison to machine learning, time series, and other algorithms. To clarify what other algorithms have been compared with the DL algorithms and through which metrics their performance has been evaluated, we explored the numerical analysis of the reviewed papers. Table 8 includes the papers that studied multiple DL techniques and compared their performances together and to the other algorithms.

The keyword “Food supply chain” also indicates the promising application of DL methods in the food industry in recent publications. Finally, the other found keywords are “Deep neural network” and “BP neural network” each of them appeared in 5 papers in the year 2021. Our primary goal was to maximize cost savings by streamlining medication procurement across this hospital network and its pharmacies. We employed advanced machine learning algorithms and predictive analytics to establish more efficient and responsive inventory management practices. The transformative potential of machine learning for supply chain managers is not a mere concept, but a reality that has been demonstrated by numerous organizations across various industries. Here, we present a selection of real-world success stories illustrating how machine learning has driven tangible improvements in supply chain operations.

There are a few other industries that are still in the early stages of utilizing DL techniques to improve supply chain performance, which might be dangerous. As mentioned earlier, RNN especially the LSTM, has been extensively used for the forecasting problem. In the logistics industry, Shankar et al. (2020) applied an LSTM network to forecast container throughput. Punia et al. (2020) presented a cross-temporal forecasting framework (CTFF) based on an LSTM network for retail supply chain demand forecasting. Koç and Türkoğlu (2021) presented a model consisting of three steps of normalization, multilayer LSTM network, and dropout-dense-regression layers to forecast the Covid-19 disease cases and the demand for medical equipment.

The recent supply-chain disruptions and demand triggered by the COVID-19 pandemic have further amplified the need for companies to develop their central-planning muscles. On the other hand, most of the papers developed a theoretical model and investigated the results through simulation and experimental analysis. The number of papers that presented practical applications of DL methods in SCM is still limited. Meanwhile, DL methods can be integrated with other technologies such as IoT and blockchain to improve the performance and integrity of supply chains. Wang (2020) proposed a classification model based on CNN, DNN, and factorization machine technology to construct a personalized tourism service system.

To maximize the benefits of supply chain optimization with AI, define the goals and expected outcomes of your AI initiatives. Establish clear key performance indicators (KPIs) to measure the success of your AI projects and align your machine learning supply chain optimization AI strategy with your overall business objectives. Before implementing AI solutions, evaluate the current state of your supply chain design to identify areas where machine learning can drive the most significant improvements.

The Growing Importance of Supply Chain Optimization

Nikolopoulos et al. (2021) also used the LSTM and many other forecasting methods to forecast the Covid-19 growth rate and compared their results. Although the LSTM is a powerful DL network, the other used methods showed better performance in this study (as discussed in Sect. 3.3.3). Weng et al. (2019a, b) employed ARIMA, BPNN, and RNN for forecasting the price of horticultural products and showed the better performance of RNN compared to the other two methods.

Analyzing historical sales data and other variables with its AI-driven models have enabled PepsiCo to make better choices regarding production levels, leading to significant cost savings and a more efficient supply chain. Transportation and global trade management software helps manufacturers plan the movement of raw materials and finished products and recommends the most cost-effective and compliant method. Although transportation managers have been using this software for many years, advances in machine learning, IoT tracking, cloud computing, and other technologies are making real-time fleet monitoring a reality. When financial data is separate from logistics data, which is separate from inventory data—that is, when every group of data lives in its own silo—achieving an end-to-end view of the entire supply chain is impossible.

Applications of Machine Learning in Supply Chain Management—A Review

The first step is a thorough analysis of all elements of the current supply chain, followed by developing or adjusting production and inventory plans that align with demand forecasts. The final phase centers on execution, during which IT and supply chain managers choose systems for inventory, warehouse, transportation management, analytics, and decision support. Supply chain managers need to constantly assess the workings of their supply chains and stay informed about new technologies and processes that will continually improve their operations.

• However, labeling mistakes or legibility problems may occur in the packaging process (Thota et al. 2020).
• For example, in a scenario in which the forecast predicted low sales of a particular SKU, planners collaborated with marketing and sales to test that prediction through demand sensing and agree on the best path forward.
• The authors used the Sample average approximation (SAA) method for solving the chance-constrained programming problem after reformulating it into a conventional chance-constrained programming problem.
• Encourage a mindset of continuous improvement and responsiveness among your employees by promoting a data-driven culture.

This assessment will make it possible for you to determine which aspects of your supply chain are most suitable for AI integration and prioritize your efforts accordingly. In recent years, supply chain disruptions have become increasingly common due to factors such as geopolitical tensions, climate change, and global health crises. These disruptions highlight the need for organizations to build resilient supply chains capable of mitigating risks and maintaining operations despite unexpected challenges.

AI can change how companies operate by providing applications that streamline planning, procurement, manufacturing, warehousing, distribution, transportation, and sales. Having the ability to monitor the whereabouts of various goods and materials while they are in transit is the foundation of an efficient supply chain, giving organizations a clear view of their activity and inventory as it moves through the supply network. ASCM is an unbiased partner, connecting companies around the world with industry experts, frameworks and global standards to transform supply chains. Some organizations believe they need to build a new tech stack to make this happen, but that can slow down the process; we believe that companies can make faster progress by leveraging their existing stack.

• This demands the collaboration, integration and sharing of information by these entities.
• Punia et al. (2020) presented a cross-temporal forecasting framework (CTFF) based on an LSTM network for retail supply chain demand forecasting.
• It combines big data (internal, external, and customer information) and advanced analytics at every step of the supply chain planning process.
• In their model, the LSTM mines information from data, and the lightGBM increases the interpretability of the model.
• Manufacturers must confront supply chain challenges, such as geopolitical instability, bilateral trade conflicts, port congestion, labor strikes, and worker shortages, that delay the offboarding of cargo shipments worldwide.

No background required, though some general knowledge of supply chain will be helpful. We asked all learners to give feedback on our instructors based on the quality of their teaching style. The Website is secured by the SSL protocol, which provides secure data transmission on the Internet. If purchasing multiple bundles, please complete your purchase before continuing with another. When you enroll in the course, you get access to all of the courses in the Specialization, and you earn a certificate when you complete the work. If you only want to read and view the course content, you can audit the course for free.

The author (Hu 2020) used DL and a fuzzy algorithm to analyze the supply chain financial credit risk. Supply chain financial credit evaluation is also beneficial for banks to control their risks and gain more profit. Zhou et al. (2020) proposed a distributed approach for supply chain financial fraud detection using a DL model. The approach discovers fraudulent financing behaviors and thus reduces supply chain losses from such behaviors. Tosida et al. (2020) applied a DL approach to classifying telematics SMEs in Indonesia based on their need for financial assistance.