AMAZON.COM, INC.- ROBOTICS Developing A Supplier Assessment Framework

BRUNSWICK CORPORATION - MERCURY MARINE

Moving From Infinite To Finite Capacity Planning

Student Team: Jian-Liang Lin – Master of Business Administration Jack Simmons – Dual MBA & MSE in Industrial and Operations Engineering Xinran (Sharon) Sun – EGL (BSE & MSE in Industrial and Operations Engineering)

Project Sponsors: John Buelow – VP of Global Operations John Cavey – Director of Global SIOP

Faculty Advisors: Oleg Gusikhin – College of Engineering Jim Omartian – Ross School of Business

Mercury Marine, a subsidiary of Brunswick Corporation, is a $3B manufacturing leader in the recreational marine boating market. They produce both engines and marine parts and accessories. The Controls & Rigging (C&R) division falls under Mercury’s Propulsion department and is a $300M business. C&R products are high margin and include switches, steering, GPS systems, Speed Display, joysticks, engine controls, and wire harnesses that connect steering controls to the engine.

Mercury’s C&R business used an unconstrained production scheduling system that functioned well when the business was right sized. However, during COVID-19, Mercury saw huge demand spikes in recreational boating. Forecasters did not predict the demand would last beyond 2022, so Mercury decided to maintain its current manufacturing footprint. It did not approve capital expenditures for facilities but did approve spending for short term capacity increase within its existing infrastructure. Despite efforts to restructure plant layouts and add headcount to existing lines, Mercury was holding a backlog 10x their previous high. In addition to the underlying capacity issue, supplier shortages were exacerbating back orders by extending lead times from weeks to months. From a plant level view, schedulers were required to make manual changes to the system-generated schedule to smooth out the fluctuating demand signals into executable production levels. Plant floor personnel in one of Mercury’s C&R departments were manually moving over $600,000 of product orders a day in 2021.

To combat these issues, the Tauber team mapped the process flow of C&R plants, interviewed stakeholders about the impact of demand changes on the S&OP process, analyzed data to narrow down the focus area, and piloted a solution. The Tauber team created a capacity constrained scheduling model with an 8-week firm horizon. It translated last minute changes and demand spikes into a predictable, efficient, and executable production schedule for the plant. The pilot resulted in 6 consecutive weeks of consistent, above average output, the longest streak of consistent production in 2021. It decreased the need for manual schedule intervention by over 84%. Ontime delivery increased by 18% in just 4 weeks.

Following the success of the pilot, Mercury plans to roll out the Tauber team’s solution to the entire global Controls & Rigging department with potential applications in the Parts and Accessories business. This is expected to deliver a cost savings of $42M by Year 3, with a potential upside of $62M from increased service levels, reduced transportation costs, increased asset utilization, and increased plant output.

CURATION FOOD, INC.

Modeling Supply Chain Risks for Agricultural Sourcing

Student Team: Brody Bushnell – Dual Juris Doctor & Master of Business Administration Dean Golan – EGL (BSE & MSE in Industrial and Operations Engineering)

Project Sponsors: Albert Bolles – Chief Executive Officer Tim Burgess – Senior Vice President of Supply Chain Jeff Korengel – Vice President of Program Management and Procurement

Faculty Advisors: Sam Aflaki – Ross School of Business Margaret Wooldridge – College of Engineering

Curation Foods, Inc., a $550M enterprise, is a natural foods company dedicated to making fresh, plant-based foods with 100% clean ingredients accessible to as many people as possible. Curation Foods is a different kind of consumer packaged goods company, with leading brands like Eat Smart® superfood salad kits and Yucatan® and Cabo Fresh® guacamole, and the premium line O Olive Oil and Vinegar®.

The Tauber team aimed to solve two challenges: (1) predicting harvest dates for broccoli when 3 weeks out from harvest and (2) quantifying the optimal oversupply percentages for broccoli plantings to maximize yearly savings and reduce costs. Significant challenges experienced by the company include lacking clear data collection strategies, weather factors affecting harvest and quality, inconsistent supply planning, and high levels of demand volatility.

The Tauber team created regression and financial models to address and solve these challenges. To address the problems of harvest date subjectivity, climate change, and record keeping, the team built the “Days to Harvest” Dashboard. This tool analyzes historical weather patterns and a series of other inputs to determine when broccoli fields will be ready to harvest 3 weeks out from anticipated harvest dates. Two prediction models based on a multiple linear regression were built for California and Arizona. The R-squared for the California and Arizona regressions were 90.4% and 87.2%, respectively. Testing of 69 plantings in the Salinas Valley confirmed that the model is 97% accurate within +/- 6 days of the actual harvest date. The dashboard also includes a centralized database with information on all harvests, ranches, growers, and crop varieties to standardize the procurement office’s data collection efforts.

The oversupply strategy ultimately produced two main recommendations using the newsvendor statistical model - one for achieving a fill rate of 98.0% and one for maintaining the historical fill rate of 96.3% for broccoli products. These two recommendations yielded an expected reduction in average yearly procurement expenditures of $627K and $2,125M respectively. Furthermore, the oversupply strategy model showed that there was great potential for increased financial savings from improving forecasting capabilities (ranging between $550K and $1.2M in yearly financial savings based on the reduction in forecasting error) and that such efforts should be pursued.

THE ESTÉE LAUDER COMPANIES INC.

Improving Ecommerce Delivery Experience Through New Inventory Positioning Capabilities

Student Team: Shreya Agrawal – Master of Science in Engineering, Industrial and Operations Engineering Claire Covner – Master of Business Administration

Project Sponsors: Sarah Bonnaud – Vice President Global Supply Chain, Online George Kuzma – Senior Vice President Global Supply Chain, North America Soo Park – Vice President, Sales Operations & Brand Planning, Online

Faculty Advisors: Mariel Lavieri – College of Engineering Joline Uichanco – Ross School of Business

The Estée Lauder Companies Inc. (ELC) are one of the world’s leading manufacturers and marketers of quality skin care, makeup, fragrance and hair care products. The company’s products are sold in approximately 150 countries and territories under brand names including: Estée Lauder, Clinique, M·A·C, La Mer, and many others. ELC started direct-to-consumer ecommerce brand sites in 1998 in the U.S. and has since continued to expand its ecommerce channels globally. ELC brands are known for high touch consumer experiences and the ELC Supply Chain team constantly focuses on elevating the brand sites’ post-purchase experience.

In the U.S., brand site orders are fulfilled in a multi-node fulfillment network managed through a third-partylogistics provider. When inventory is not positioned effectively across the fulfillment nodes, it impacts the brands’ speed of delivery and, in some cases, leads to split orders.

The Tauber team was tasked with diagnosing root cause issues contributing to split orders and non-optimal shipments, assessing current inventory positioning technology, and piloting new solutions to maximize orders shipped from fulfillment centers closest to the consumer and minimize split orders.

The Tauber team approached this problem by first conducting a current state assessment to identify opportunities for improvement and then validated several hypotheses on root causes, using supporting data analysis. From there, based on expected impact, the team focused its analysis on an assessment of current forecasting processes and tools and recommendation of short-term tactics ELC should use to drive improvement given the current state of inventory positioning. Overall, the Tauber team recommended and initiated ELC’s longer term movement toward additional inventory positioning technology capabilities and new strategies for short-term responsiveness to current positioning. These changes are set to yield significant time savings, cost savings, and improvement in process inefficiencies along with benefiting additional brands within the ELC portfolio and other global markets that are also moving to multi-node distribution networks.