Supply Chain

The LGO program has many ways to focus your MBA on supply chain management. There is a defined track within Civil and Environmental Engineering for students who want to dive deep into supply chain coursework. Students then have multiple companies who regularly offer internship projects on global supply networks, which allow students interested in supply chain management to work in the field before taking a leadership role after graduation.

Virtual Pooling Approximation Using Longest Path Network Optimization

Kevin Schell (LGO ’18)

Company: Caterpillar
Location: Aurora, IL

Problem: Caterpillar sells its products through a network of independently-owned dealers which own their inventory. Given long factory lead times, highly variable demand across geographies, and complex configurations, dealers are forced to maintain high inventory on their sites in order to keep service levels up. Kevin’s project aimed at exploring pooling arrangements among dealers to address these challenges, while taking into account the dealers’ reluctance to unilaterally give up equipment they might eventually sell.

Virtual Pooling Approximation Using Longest Path Network Optimization
Virtual Pooling Approximation Using Longest Path Network Optimization

Approach: Kevin proposed a model which Caterpillar dealers would maintain physical inventory in multiple locations and use a network swapping mechanism to create a virtual pool from which various dealerships could make and fulfill sales. He defined a commercial network model to simulate equipment ordering, inventory management, and sales across a sample dealer network, and ran the model to simulate results with and without network swapping implemented.

Impact: The baseline simulation results from Kevin’s model for a single class of Caterpillar vehicle suggest that network swapping could reduce inventory by over 12% and reduce customer back orders by over 17%. The net present value of the Caterpillar dealer network could be increased by over $3M. Kevin’s thesis addressed these results and outlined how this network swapping model could be further improved.

Shipping Pricing for Build-to-Order Products

Carrie Beyer (LGO ‘17)

Company: Dell
Location: Austin, TX

MBA supply chain internship Dell
Carrie’s logic in building her optimization model

Problem: has uniform shipping pricing across Build To Order (BTO) notebook products. In October 2016, Dell launched a program to give customers more information and better shipping method choices. They asked Carrie to develop an optimization model that would recommend shipping prices to maximize the value for the company across three areas: customer experience, profitability, and working capital.

Approach: Carrie identified key variables within each area that would impact delivery. She concluded that price elasticity was an important variable for which little data was available. To collect this data, Carrie designed a price elasticity experiment to forecast demand at different price points. She used her results to develop a linear program optimization model to calculate optimal prices for each level of service.

Impact: Carrie’s model indicated opportunity to increase delivery value by 28%. To implement the project, Dell has scheduled necessary IT changes.

Amazon Prime Pantry Operations

Nupur Dokras (LGO ’17)

Company: Amazon
Location: Seattle, WA

Problem: Amazon’s Prime Pantry has recently shifted to provide more selection of small, everyday products. Amazon wanted a way to identify which products they should include in the Pantry line that would both maximize efficiency in delivery and minimize the products’ costs.

MBA supply chain internship Prime Pantry
Nupur’s modeling results

Approach: To identify which products to include, Nupur studied historical data on inbound profile, cubic volume, and demand. She created a dynamic model in which Amazon can enter product characteristics and view projected operational costs throughout the Prime Pantry network. The model uses minimum pallet quantities, shelf life, vendor lead times, and demand to determine how much inventory will be in the fulfillment center at a given time. Her analysis showed that the picking process costed the most of all steps in fulfillment. She conducted testing on this step, including splitting the process paths for lower and higher units.

Impact: Nupur’s dynamic model provided a cost breakdown and recommendations for order and storing inventory. Retail teams and Pantry sites now use this tool to manage inventory on a more granular level. Her new fulfillment solutions have been implemented across the Pantry network and has been successful in increasing the weekly profit for the business while decreasing variable cost per unit.