Cambridge, MA – The MIT Supply Chain Management Master’s Program has recognized thirty-five exceptional students from eight renowned undergraduate programs specializing in supply chain management and engineering across the United States.

Presented annually, the MIT Supply Chain Excellence Awards honor undergraduate students who have demonstrated outstanding talent in supply chain management or industrial engineering. These students originate from institutions that have partnered with the MIT Center for Transportation and Logistics’s Supply Chain Management master’s program to expand opportunities for graduate study and advance the field of supply chain and logistics.

In this year’s awards, the MIT SCM Master’s Program has provided over $900,000 in fellowship funding to 35 deserving recipients. These students come from respected schools like Arizona State University, University of Illinois Urbana-Champaign, Lehigh University, Michigan State University, Monterrey Institute of Technology and Higher Education (Mexico), Penn State University, Purdue University, and Texas A&M University.

Recipients can use their awards by applying to the SCM program after gaining two to five years of professional experience post-graduation. The fellowship funds can be applied towards tuition fees for the SCM master’s program at MIT or at MIT Supply Chain and Logistics Excellence (SCALE) network centers in Spain, Malaysia, Luxembourg, or China.

For more information about the MIT Supply Chain Excellence Awards, please visit here.

___________________________

About the MIT Center for Transportation & Logistics

Founded in 1973, MIT CTL is one of the world’s leading supply chain education and research centers. MIT CTL coordinates more than 100 supply chain research efforts across the MIT campus and around the globe. The center also educates students and corporate leaders in the essential principles of supply chain management and helps organizations to increase productivity and improve their environmental performance.

About the MIT Supply Chain Management Master’s Program (MIT SCM)

Founded in 1998 by the MIT Center for Transportation & Logistics (MIT CTL), MIT SCM attracts a diverse group of talented and motivated students from across the globe. Students work directly with researchers and industry experts on complex and challenging problems in all aspects of supply chain management. MIT SCM students propel their classroom and laboratory learning straight into industry. They graduate from our programs as thought leaders ready to engage in an international, highly competitive marketplace.

Supply Chain Excellence Contact: Kate Padilla kpadilla@mit.edu

Cambridge, MA – The AWESOME award represents a significant commitment by the MIT Supply Chain Management Master’s Program, the MIT Center for Transportation & Logistics, and AWESOME (Achieving Women’s Excellence in Supply Chain Operations, Management, and Education) to encourage students to prepare for and perform successfully in supply chain leadership roles. This fellowship was awarded to two students each year: one from the residential cohort and one from the blended.

Editor’s Note: The SCM capstone Case Fill Rate Prediction was authored by Madeleine Lee and Kamran Iqbal Siddiqui, and supervised by Dr. Elenna Dugundji (elenna_d@mit.edu) and Dr. Thomas Koch (thakoch@mit.edu). For more information on the capstone, please contact the thesis supervisors.

The fast-moving consumer goods (FMCG) industry, currently valued at $10 trillion, is poised for exponential growth, projected to reach $15 trillion by 2025. However, the industry’s complex supply chains and unpredictable demand patterns have been further complicated by recent disruptions such as the COVID-19 pandemic, the Suez Canal blockade, and port congestion. These challenges have highlighted the vulnerability of the FMCG sector, leading to unmet customer demands and jeopardizing business resilience.

To address these issues, our capstone project set out to identify the key factors contributing to low case fill rates (CFR) in an FMCG company and develop predictive models to improve future CFR. By harnessing data-driven insights, this project offers a transformative approach to managing CFR, ultimately enhancing sales, customer loyalty, and overall resilience within the FMCG landscape.

Common struggles

Consumer product companies, with diverse product portfolios, face the common struggle of maintaining optimal CFR in a dynamic business environment. CFR, calculated by dividing total shipments by total customer orders, serves as an indicator of a company’s ability to fulfill customer demands. A decline in CFR can result in lost sales, eroded customer loyalty, and even potential contract breaches. As little as a 1% drop in sales can translate into millions in lost net profit margin, underscoring the criticality of maintaining an optimal CFR.

Our project followed a robust three-phase methodology: business understanding, modeling, and validation. Extensive datasets, including sales transactions, customer purchase history, inventory levels, and manufacturing plans, were compiled to gain a holistic understanding of business operations and customer behavior. Rigorous data preprocessing techniques were employed to ensure data integrity, integrate diverse datatypes, and enhance data quality. And descriptive analytics, such as ACF (autocorrelation function) and PACF (partial autocorrelation function), were applied to identify autocorrelation in the time series dataset.

In the modeling phase, various techniques were employed to uncover patterns and forecast future CFR. Time series analysis enabled the isolation of the impact of single and multiple variables on CFR. Statistical methods, including decision tree matrices, were utilized to identify major drivers of low CFR. Two distinct approaches were taken: a hybrid model to predict cut quantities and, to forecast inventory availability and order quantity, advanced deep learning techniques, including XGBoost, LSTM, and multistep LSTM.

Improving forecast accuracy

The project yielded valuable insights into the factors influencing CFR. It was found that forecast accuracy and demand variability were the primary drivers impacting CFR. Inaccuracies between predicted and actual customer demands significantly influenced the CFR, underscoring the importance of improving forecast accuracy and maintaining adequate inventory levels.

The hybrid model, incorporating logistic regression, random forest regression, and support vector machine (SVM), demonstrated impressive precision and accuracy in predicting cut quantities and CFR. In the second approach, advanced time series machine learning models like XGBoost, LSTM, and multi-step-LSTM showed potential for short-term forecasting. While long-term predictions posed challenges, these models provided valuable insights for inventory optimization.

The project’s findings hold immense significance for the FMCG industry. By identifying forecast error and demand variability as critical factors impacting CFR, the project highlights the need for improved forecast accuracy and inventory optimization. Integrating exogenous factors, such as promotions and market indices, into forecasting models can further enhance accuracy and reliability.

Though challenges remain, such as inventory variability and irregular order patterns, the project lays a strong foundation for future research. Employing a combination of machine learning and deep learning techniques, exploring reinforcement learning, and incorporating additional data inputs like promotional activities and competitor pricing can lead to superior predictive accuracy. By adopting these data-driven insights, FMCG companies can proactively mitigate stockouts, enhance CFR, and thrive in an ever-changing business landscape.

Supply Chain Management Review

Every year, approximately 80 students in the MIT Center for Transportation & Logistics’s (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects.

These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.

Made possible by a grant from the UPS Foundation, the UPS Fellowship continues its mission to champion outstanding students with financial support of two exceptional students, one incoming MIT Master’s student and one MIT PhD student pursuing scholarship relating to logistics, freight transportation, supply chain management, or a related topic. A continuation of a program started in 1983, the UPS Fellowship aims to recognize and reward excellence in these fields, and selections are awarded solely on the basis of merit.

This year’s fellowship recipients are:

“The UPS Fellowships exemplify MIT CTL’s dedication to infusing innovation into real-world applications, upholding the highest standards of academic inquiry,” said Chris Caplice, Executive Director of MIT CTL. “These fellowships, with the generous backing of the UPS Foundation, stand as indispensable assets in nurturing talents such as Erin and Steven. Their contributions will help to shape the future landscape of the supply chain industry.”

Please join us in congratulating Erin and Steven!

By Elenna Dugundji, Andres Ayala, Ria Verma, and Thomas Koch 

Of the numerous AI applications in supply chain management, supplier selection, risk resilience, and contract negotiation are often cited as offering the most potential for generative AI. The need for high volumes of text and data makes these areas particularly suitable for such applications. However, many organizations struggle to overcome the complexities of integrating generative AI into mainstream operations, one reason that projects often fail.

A pilot project at the MIT Center for Transportation & Logistics (MIT CTL) aims to overcome these complexities and demonstrate how targeted, real-world applications of generative AI can be successfully implemented in the procurement function. The project team is developing a chatbot for a leading pharmaceutical company with a direct and indirect annual spend of over $35 billion. The bot will help category managers negotiate more effectively with suppliers by providing comprehensive information on key questions like how prices are trending for specific materials.

Addressing data issues

Category managers rely on a multitude of data sources, including spend analyses, bills of materials (BOM), and other industry-specific information, to devise effective negotiation strategies. However, existing sources of this data are often underutilized because they are not easy to access or are relatively limited. For example, ERP systems represent a key source, but the number of systems and their complexity can be problematic for users. Also, current methods for data retrieval, such as visualization dashboards, often require extensive navigation and a steep learning curve for newcomers. Category managers can turn to data scientists for ad-hoc analyses. Still, these specialized staff may be in short supply, and relying on them too much creates bottlenecks that impede the information-gathering process.

With these difficulties in mind, the pilot’s goal is to augment (increase efficiency) and automate (reduce time) the collection and collation of information that category managers need when negotiating with suppliers. While the necessity of data scientists would not be reduced, generative AI could “place a data scientist in the pockets” of every procurement professional in this application area. As a result, the technology would increase managers’ efficiency, democratize data access, and foster data-driven decisions in supplier negotiations. These benefits could be realized regardless of the category managers’ technical prowess or experience level, allowing procurement professionals to focus more on strategic aspects of their jobs.

Choosing the best approach

A defining feature of generative AI is its ability to provide coherent and contextually relevant responses and flexible phrasing options that allow questions to be framed differently. For example, the technology can translate free-form text into SQL or another data-querying language to extract information, thus extending the scope of inquiries beyond pre-defined questions. This capability differentiates the technology from traditional chatbots based on fixed question-and-answer pairs. Furthermore, generative AI can empower users to perform complex tasks like generating graphs and conducting statistical analyses without requiring a coding background. When integrated with more sophisticated models, these tools can even undertake advanced tasks such as predictive and prescriptive analytics, showcasing their versatility and depth in creating new insights from the available data.

Given these capabilities, it is vitally important that the new chatbot is designed to respond to the types of questions the pharmaceutical company’s category managers typically ask in preparation for negotiations with suppliers.

The insights the company’s procurers look for vary in complexity. We initially categorized these questions into separate use cases and organized them into three buckets. This categorization is crucial as each question type can benefit from different models (see Figure 1).

A retrieval-augmented-generation (RAG) approach was employed with the company’s actual data. The RAG approach involves retrieving data from a knowledge database relevant to a question and providing it as context to large language models (LLMs) that generate a response. The model was deployed using AzureOpenAI’s LLM within a secure environment. The RAG method can be advantageous for reducing inaccuracies or “hallucinations” primarily because it prioritizes fetching information from the existing knowledge base, ensuring that the content is anchored to retrieved, reliable texts.

An important question for the project team is whether or not it is better to use an off-the-shelf AI solution to meet these various demands rather than developing a solution in-house. A bespoke chatbot can avoid the time and cost of developing a tailored solution. However, the team opted for an in-house solution for several reasons.

Generative AI is not protected against the well-established “garbage in, garbage out” or GIGO effect, which underscores the importance of data quality and modeling. The complexity of the models and data architecture significantly impacts the resulting output. Hence, the benefit of designing a tailored model while still using a proprietary LLM is the increased flexibility this affords data scientists to swiftly iterate, test various models, and deploy solutions in secure, private environments. Another advantage of the in-house option is that it can be tailored to specific organizational needs while keeping costs and the need for computationally skilled persons manageable. Furthermore, given the experimental nature of this field, piloting a proof of concept first can help foster trust from relevant stakeholders and make it easier to secure investment funds.

The project’s “low-hanging fruit” was to address Type 1 questions first (see Figure 1). Preliminary iterations utilized a LangChain SQL Database agent (LangChain is a Python library that offers tailored LLM applications that can be deployed for different tasks using different agents). Much higher accuracy was achieved when using careful prompt engineering (designing the input to produce an optimal output), such as formatting the free-form text intentionally to replicate query language.

This approach poses two critical questions: how can the likelihood of user error be accounted for, and how can category managers be protected from potentially inaccurate information when questions are posed ambiguously?’

Figure 1: Category manager questions

One strategy is to encourage category managers to learn and embrace pseudocode—a mix of plain language and coding syntax that explains how a program should work—without using actual programming language. Even without a technical background, using words like filter and aggregate or breaking down complex queries into smaller sentences first can greatly increase the tool’s accuracy.

Another approach focuses on refining the user interface. Implementing a dropdown menu can guide users away from entering free-form text where this type of input is not ideal for the model. A help area that clarifies users’ objectives before inputting free-form text can be included. In addition, an instruction-tuning facility at the back end of the application can guide the agent in answering a specific category of questions in a purposely directed way.

Next steps

Following a review of the preliminary chatbot version by the sponsor company’s CPO, the plan is to refine the model and develop a comprehensive roadmap to scale it further. A particularly promising avenue involves using a graph database or knowledge graph instead of a relational database to establish connections between BOMs and spend data, unlocking more profound insights geared toward addressing more complex questions. This refinement represents a significant opportunity to enhance the procurement organization’s analytical capabilities.

We also intend to research and develop a roadmap to facilitate the full-scale deployment of a fully accessible chatbot. This will involve outlining the roles and responsibilities of the functional groups involved in this process.

Significant challenges must be overcome before the chatbot becomes integral to the company’s procurement operations. These include data quality and access issues, API permissions for security, difficulties of latency and accuracy when using relational databases for the LLM, and the reliability of deployed applications.

However, the chatbot has the potential to deliver substantial benefits. Also, the project could set the stage for steady, transformative progress in advanced AI and establish a new benchmark for efficiency and innovation in procurement.

Supply Chain Management Review

By Jorge Oliver and Aaron Zheng · December 13, 2023

Editor’s Note: The SCM thesis Development and Evaluation of Market-Based Routing Guide Strategy was authored by Jorge Oliver and Aaron Zheng, and supervised by Dr. Chris Caplice (caplice@mit.edu). For more information on the research, please contact the thesis supervisor.

The truckload (TL) transportation market in the United States is large, fragmented, and highly competitive. Shippers manage their carriers using a routing guide within their Transportation Management System (TMS). The routing guide is the bridge between a shipper’s strategic procurement (usually through a reverse auction or Request for Proposal process) and their tactical execution. It specifies which carrier is the primary for each lane.

Our research examined how routing guides perform across different types of lanes. Specifically, we assessed the “macro-market” and “micro-shipper” effects. The macro-market perspective looks at each lane nationally, considering shipping data across 3 million lanes from TMC, a division of global logistics company C.H. Robinson, and the company’s Procure IQ tool, while the micro-shipper perspective considers only an individual shipper’s volume. For each perspective, a lane can be classified as being in one of four quadrants: Balanced (where there is high volume in both directions), Headhaul (where there is high volume from origin to destination but low volume in the other direction), Backhaul (where there is low volume from origin to destination but high volume in the other direction), and, finally, Sparse (where there is low volume in both directions). The only differences between the macro and micro perspectives are the specific target number for high versus low volume levels and what truckload volumes to consider (all shippers versus just one shipper).

Figure 1: Shipping lane quadrants

Before exploring routing guide performance, however, we needed to establish whether the macro-market classifications were stable over time. Our analysis showed that these lane classifications were very stable at the macro level. We found that 78% of the lanes, defined as a Key Market Area (KMA) to KMA pair, did not change quadrants over the eight years (2015–2022) of market data.

Then, we assessed the routing guide performance for lanes based on the primary carrier acceptance rate. We found no significant variance in performance across the four macro-market categories. However, we discovered a significant difference when examining lanes at the micro-shipper level. Balanced and Headhaul lanes exhibited at least an 8% higher primary carrier acceptance rate than Backhaul and Sparse lanes. This suggests that the shipper’s freight flows influence carrier behavior more than the broader macro-market flows.

This result suggested an opportunity to leverage the macro-market level to improve the routing guide performance for those lanes that are low volume at the micro-shipper level but high volume at the macro-market level. While these lanes only handle about 9% of the volume, they represent over half (about 53%) of the lanes a shipper manages.

Based on our findings, we developed a strategic procurement framework that classifies lanes into four potential relationships that should be part of every shipper’s procurement portfolio.

1. Lanes that are Balanced at the micro-shipper level are characterized as having high and predictable volumes in both directions and should be considered for private/dedicated fleet or contracted capacity.

2. Lanes classified as Headhaul at the micro-shipper level are primary candidates for traditional one-way over-the-road contracts.

3. Lanes that have low volume at the micro-shipper level (Backhaul or Sparse) but are high volume (Balanced and Headhaul) at the macro-market level are candidates for a structured spot pricing strategy where the shipper and carrier agree to set the price per load dynamically based on a mutually agreed upon 3rd party index pricing.

4. Shippers should consider following a traditional market spot approach instead of investing resources in establishing contracts for the remaining lanes.

By applying this strategic procurement framework that considers both a lane’s macro-market and micro-shipper characteristics, shippers can leverage a wider portfolio of relationships that improves their overall routing guide performance while reducing the required effort.
Every year, approximately 80 students in the MIT Center for Transportation & Logistics’ (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects. These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.

Supply Chain Management Review

Cambridge, MA – MIT’s Supply Chain Management master’s program has been ranked the #1 SCM master’s in the world by Quacquarelli Symonds (QS) for the fourth consecutive year. For its 2024 rankings, QS evaluated the SCM master’s programs on the basis of  employability, alumni outcomes, value for money, thought leadership, as well as reputation among industry professionals and academics. MIT SCM Master’s Program received an overall score of 100 out of 100 on the QS Value. This ranking relies on input from the QS Global Employer Survey, where thousands of employers  identify their preferred schools for recruitment. 

The SCM program at MIT offers a distinctive fusion of executive leadership training and an intensive core curriculum, placing a strong emphasis on the development of analytical and technical competencies.The program combines executive leadership training with an intensive, practical core curriculum focused on building analytical and technical knowledge. In just ten months of cohort-based, full-time on-campus study, students develop critical  reasoningskills top employers look for. Offered through MIT’s Center for Transportation and Logistics, with cross-registration opportunities at the Sloan School of Management, MIT’s SCM program leads to an engineering degree and offers stellar post-graduate outcomes. A hybrid Blended program option gives students who’ve completed the online MITx MicroMasters in SCM the opportunity to earn their SCM masters in just one semester (five months) of full-time study on campus.

___________________________

Furthermore, it’s worth noting that the MIT School of Engineering, the home of the MIT SCM program, holds the top spot as the #1 engineering school, as reported by US News and World Reports. MIT SCM students have the opportunity to enroll in courses from various departments within the School of Engineering as part of their academic journey (source: https://www.usnews.com/best-graduate-schools/top-engineering-schools/eng-rankings).

Additionally, the Massachusetts Institute of Technology has once again secured the coveted top position as the world’s number one university, according to the prestigious QS Rankings (source: https://www.topuniversities.com/university-rankings/world-university-rankings/2024).

About the MIT Supply Chain Management Master’s Program (MIT SCM)

Founded in 1998 by the MIT Center for Transportation & Logistics (MIT CTL), MIT SCM attracts a diverse group of talented and motivated students from across the globe. Students work directly with researchers and industry experts on complex and challenging problems in all aspects of supply chain management. MIT SCM students propel their classroom and laboratory learning straight into industry. They graduate from our programs as thought leaders ready to engage in an international, highly competitive marketplace.

Media Contact: Lisa Kim Lisahuh@mit.edu

By Geoffrey Allen and Shoichi Ishida 

Editor’s Note: The SCM thesis B2B Omnichannel Network Design and Inventory Positioning was authored by Geoffrey Allen and Shoichi Ishida and supervised by Dr. Eva Ponce (eponce@mit.edu). For more information on the research, please contact the thesis supervisor.

The B2B foodservice distribution industry in the United States is projected to witness substantial growth over the next five years, but like many B2B industries, it is just starting to experiment with omnichannel fulfillment. Some companies have taken the plunge and borrowed from the B2C playbook, offering fulfillment through stores in addition to the more traditional direct delivery to their customers, while others are doubling down on delivery with new and more flexible delivery options. In either case, the challenge lies in redesigning supply chain networks to accommodate increased complexity and flexibility without radically inflating costs. Companies must consider factors such as inventory positioning, shrink, lead time, and more when determining how many products should be transported between suppliers, warehouses, and customers. Unfortunately, while B2C omnichannel network design is a popular research topic, B2B applications to date make up only 4% of reviewed studies and pose unique challenges.

Why don’t existing approaches or popular software packages work for us?

Implementing familiar software solutions to address these challenges may seem like a straightforward decision, but this comes with its fair share of challenges. Let’s explore three key issues.

Firstly, complicated and interrelated omnichannel options can stress existing software packages to the breaking point. Common network design approaches assume a hierarchical and sequential flow of goods from supplier to customer. However, true omnichannel designs, where all order channels for all items can potentially be fulfilled across all delivery options, can overwhelm the design user interface for common packages due to the combinations involved. This is especially true for companies with a large number of items, suppliers, and/or customers. More exotic, but potentially cost-saving, designs involving options like lateral transshipments between nodes exacerbate the problem further or may not be supported at all.

Secondly, inventory plays a crucial role in the overall supply chain strategy, yet it’s not always adequately addressed by existing software. Many packages handle the nonlinear nature of inventory equations by solving for inventory positioning separate from network design. However, this frequently leaves money on the table. Items with high inventory holding costs and/or variable demand are typically held at too many locations with sequential solve approaches, while items with the opposite characteristics could be further decentralized. Some companies attempt to get around this trade-off by iterating between network design and inventory modeling with the hope of converging towards a reasonable answer, but this then increases the complexity of the overall effort.

Lastly, solve time can be a limiting factor. As models grow more complex to handle all the flexibility required, the time required to solve problems increases significantly. This can make the intelligence from these efforts less useful to decision-makers—or less realistic by requiring unpalatable aggregations to model realistic item, supplier, and customer counts.

Is there a better way?

To tackle these challenges, we developed a specialized mathematical optimization proof-of-concept model tailored specifically for a large B2B food distribution company in the United States looking to expand omnichannel distribution options. Putting it to the test with real company data, this model can solve for a global optimal solution for both network design and inventory positioning simultaneously, at close to 10 times the speed of alternative benchmarks.

Our tailored and integrated approach has shown a significant 7.6% reduction in total supply chain costs, encompassing product cost, transportation cost, warehouse handling, inventory holding, and omnichannel expenses like cross-docking or direct ship. Across the items examined, suggested inventory holding costs reduced by up to 50%.

The secret sauce lies in our custom algorithm, which leverages new capabilities in commercial solvers to effectively handle quadratic equations. We reformulate the nonlinear inventory equations into a form that allows for simultaneous solutions, and then improve the solve time using outer approximation techniques. This enables us to efficiently and quickly solve complex, real-world network optimization problems without long delays or sacrifices in realism. While our initial proof-of-concept assumed the current network assets, this model could easily be expanded to include facility open/close decisions as well.

Omnichannel network design is a complex endeavor, but with the fusion of the latest solver technology, new advanced mathematical algorithms, and a solid foundation of supply chain theory, companies can squeeze new life and greater efficiency out of their existing networks. In industries like foodservice distribution, with razor-thin margins of 2% or less, this can be a critical differentiator between success and failure.

Every year, approximately 80 students in the MIT Center for Transportation & Logistics’s (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects.

These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.

Supply Chain Management Review

By Andrew Mohn and Adriele Pradi 

Editor’s Note: The SCM thesis Leveraging Simulation-Based Optimization to Generate Optimal Transportation Plans in the Real World was authored by Andrew Mohn and Adriele Pradi and supervised by Dr. Ilya Jackson (ilyajack@mit.edu). For more information on the research, please contact the thesis supervisor.

Transportation planning for the furniture and home goods industry presents a complex puzzle. Online retailers, like the company we’re about to uncover, face the challenge of efficiently moving products from warehouses to customers’ doorsteps. However, traditional planning methods often fail to account for the inherent randomness and variation in real-world operations.

In our research, we explore the transformative journey of a large online retailer as it embraces the power of probability in middle-mile transportation planning. Their approach relied on point forecasts, not including the unpredictability of real-world operations. Consequently, disruptions, suboptimal outcomes, and increased costs were part of their logistics network, challenging their promise of timely deliveries.

Our focus was on developing a solution that leveraged probability distributions and stochastic inputs to create robust transportation plans. The key challenge was to incorporate these methods effectively, considering the complexity of implementation and the need for computational resources. To tackle this problem, we developed a simulation model within a scenario-based framework. This approach involved sampling uncertain inputs’ distributions and repeatedly solving the optimization problem with realized values. By doing so, we aimed to evaluate the model’s performance, identify blind spots, and assess improvements in plan optimization.

Embracing the unpredictable for optimal delivery

The company’s current system relies on point forecasts, which do not adequately consider the randomness and variation inherent in real-world operations. The research was motivated by the need to develop a solution that more accurately accounts for the probability distributions of inputs. Techniques such as stochastic and robust optimization, simulation and scenario testing, and reinforcement learning have been shown to address this issue effectively. However, implementing these methods at scale poses challenges in terms of expertise and computational resources.

To address these challenges, we focused on developing a simulation model using a scenario-based framework. The goal was to provide a practical approach to incorporating probability distributions into transportation planning. By leveraging this approach, our project aims to analyze and mitigate the impact of uncertainties in two primary sources of variation: facility operations and driving time between facilities.

The objective of our project is to improve plan performance and minimize disruptions by evaluating the simulation model’s effectiveness in comparison to the existing approach. The central metric of evaluation is the company’s ability to fulfill the customer promise, which measures its consistency in meeting delivery timelines.

Uncovering logistics uncertainty through data and simulation

Simulation results demonstrated differences between the current transportation plan outcomes and those generated by the scenario-based model. Certain lanes showed higher yard dwell times in the simulation compared to the service level agreement-based plan, while others exhibited the opposite trend. This suggests that service level agreements may be overly optimistic or conservative for specific lanes. Our research showcased the capability of stochastic optimization models to capture the inherent variability in transportation time and dwell, enabling more robust transportation plans that better reflect real-world conditions.

While our study highlighted the benefits of stochastic optimization, it also identified certain limitations. The accuracy of input data, specifically estimated distributions of transportation time and dwell, plays a crucial role in determining result quality. To address this, future research should focus on improving these estimates by incorporating real-time data and leveraging advanced analytics techniques. Moreover, the computational complexity associated with stochastic optimization models may pose challenges in their practical implementation for large-scale transportation networks. Exploring avenues to enhance the efficiency of these models would render them more accessible and feasible for widespread adoption.

Every year, approximately 80 students in the MIT Center for Transportation & Logistics’ (MIT CTL) Master of Supply Chain Management (SCM) program complete approximately 45 one-year research projects.

These students are early-career business professionals from multiple countries, with two to 10 years of experience in the industry. Most of the research projects are chosen, sponsored by, and carried out in collaboration with multinational corporations. Joint teams that include MIT SCM students and MIT CTL faculty work on real-world problems. In this series, they summarize a selection of the latest SCM research.

Supply Chain Management Review