Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 85-86). / Amazon is able to deliver millions of packages to customers every day through its Fulfillment Center (FC) network that is powered by miles of material handling equipment (MHE) such as conveyor belts. Unfortunately, this reliance on MHE means that failures could cripple an entire FC. The exceptionally high stock-out cost associated with equipment failure means spare parts must always available when required. This is made difficult as Amazon does not hold any central repository of inventory at present - all inventory is held at a site-level. Unfortunately, FCs have to stock more inventory than required due to unpredictable failures, long lead times from suppliers, and no standard work processes for site-to-site transfers. However, if Amazon is able to pool its spares across multiple FCs, it has an opportunity to reduce the spares kept across the entire FC network, position itself to better respond to catastrophic failures, and consolidate interfaces with suppliers. The goal of this thesis is to identify the inventory model and network design that would maximize parts availability while minimizing cost. Additionally, an implementation roadmap will be developed to outline how such a system (e.g. hub locations, logistic channels etc.) can be developed. This thesis concludes by proposing potential extensions of the work conducted in this thesis to improve the practicality and financial impact of the proposed network and inventory model. / by Nigel Goh Min Feng. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Aeronautics and Astronautics, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Page 61 blank. Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 57-58). / Thermal spray is a manufacturing process where melted particles are sprayed onto a surface to build up thickness. It is used extensively in the aerospace industry to improve and repair part surfaces, extending the useful life of expensive components. Because thermal spray is a special process, current quality practices require destructive testing. Because testing using production parts is expensive, quality management is commonly done through the use of representative test coupons. However, the coupon process is both financially expensive and operationally inefficient. Connecticut Rotating Parts (CTRP) is an FAA Part 145 repair station specializing in rotating hardware. Thermal spray is used at multiple stages during the repair process so the continued operation of its spray booths are critical to meeting delivery dates. Currently, CTRP runs weekly coupons for every material and spray booth combination. Each test cycle is at least 24 hours during which no parts can be sprayed. The objective of this thesis is to use CTRP as a benchmark to investigate thermal spray quality related issues in order to evaluate best practice quality control methods. Specifically, this project evaluated a camera system that monitors the state of the particles prior to substrate contact as an indirect measure of buildup quality. An analysis of CTRP's historical coupon and production performance showed very few failures. The failures that did occur were most likely the result of isolated deviations rather than systemic faults. Testing of the camera system was unable to conclusively establish the parameters needed for regular plume and equipment monitoring. These findings suggest that existing process controls are very capable of producing high quality coatings even in high turnover shops like CTRP and that weekly testing may be overly conservative. However, non-destructive testing methods do not yet exist that can sufficiently replace the utility of representative coupons. / by Lingmiao Wang./ / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Aeronautics and Astronautics

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 53-54). / In the face of e-commerce growth and rising customer delivery expectations, companies must adapt to meet shorter contract shipping requirements with existing infrastructures. The "Amazon effect", an evolution resulting in increased online shopping and direct-to-consumer order fulfillment, is reverberating through the retail industry and requiring manufacturers to evaluate their existing supply chain networks to meet two- and one-day shipping from their distribution centers (DC). This thesis evaluates speed and execution improvements using automated storage and retrieval systems (ASRS) in DCs. Adopting ASRS can provide the essential capability upgrades to reduce material processing time and meet direct-to-consumer deliveries. The ASRS analysis reviews current DC metrics, future throughput and inventory requirements, comparisons of ASRS technologies utilized today, expected impact of ASRS inside of an existing DC, and sizing and selection of an ASRS. Analysis of the DC evaluated showed that the primary cause of delays in shipping time was due to high variability in task completion time, rather than a high average completion time, causing extended wait times to propagate throughout the DC. While methods to reduce task time variation can be implemented, a warehouse logic upgrade would allow for real-time sequencing to get products shipped in the correct order based on factors like shipping method, customer, or priority. Implementation of ASRS in the DC evaluated could decrease processing time in storage and retrieval by 67% and total processing time through the DC by 37% due to ideal sequencing, diminished downstream variability, and reduced work in progress. The payback period for ASRS is projected to be 4 to 5 years.. / by Adriane Turner. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from PDF version of thesis. / Includes bibliographical references (pages 89-90). / Variations and the negatives effects it causes on production capacity and planning are topics of significant interest in the manufacturing communities. This research investigates the hypothesis that, when operating in a highly constrained environment, capacity can be gained by reducing the variations within the system. This study tests this hypothesis through simulation, data analysis, and controlled testing on the variations responsible for limiting capacity at Vektek LLC. The variability in lead times, quality, batch ordering, and demand forecasting contributes to the Bullwhip Effect. This increase in variability will cause excessive inventory, overtime costs, unacceptable service levels, high production costs, and large lead times. This research reduces these variabilities by isolating each cause of variability and placing standard work around it, such as SOPs. Once isolated and controlled, variations were methodically reduced, and significant capacity was gained. The research results show us that the overall variations were reduced by 26.5%. Due to this: overtime costs were reduced, late shipments were reduced by 40.0%, WIP inventories were reduced by 38.0%, and lead times were reduced roughly 22%. The total monetary value saved is estimated to be $988k and the total capacity gained was 30.8%. These results provide an initial validation that reducing the variations will increase the capacity. / by Michael C. Ross. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (page 41). / This project set out to create a framework for conducting a cost benefit analysis of part proliferation looking into first, second and third order effects of part specialization within the entire Caterpillar enterprise. This project builds on previous internal efforts to reduce complexity by evaluating the impact of increasing part count on design, procurement, inventory, production and both internal and external quality. Part proliferation occurs as parts are designed or redesigned to increase safety, comply with changing regulatory rules, improve profitability, serve niche customer demands and increase percentage of industry sales (PINS). The main driver for creating unique parts instead of common components comes from the incentivization to optimize designs for individual models and applications with a relatively narrow perspective on the cost function underlying parts proliferation. / Caterpillar factories assemble final products from unique components sourced from both internal and external suppliers. Part proliferation increases inventory and requires design and upkeep actions to create and produce the new product. Many of the challenges associated with proliferation are hidden or poorly understood as they involve factory and quality efficiencies which tend to be aggregated at a high level or dealt with as a one-time issue. Other benefits such as inventory reduction are clearer but were analyzed by this project to understand the total impact of a unique part to the system. This project focused on decreasing the proliferation of axle options within the Medium Wheel Loader (MWL) and Large Wheel Loader (LWL) product families, with the outcome of creating a generalized framework for use throughout the enterprise on any product family. / The study found that the impact of including the areas of operational inefficiencies, internal quality, and external quality, when adding an axle configuration increased the costing analysis by 64%. The details of this analysis are presented in the following dissertation. / by Paige Denise Youngerman. / S.M. / M.B.A. / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering / M.B.A. Massachusetts Institute of Technology, Sloan School of Management

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / "May 2020." Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 110-115). / Over the past one hundred years, maintenance concepts have evolved from a simple "fix when broken" approach to advanced prognostic methods used today that leverage large amounts of historical, operational, and primary sensor data to predict when and how failures will occur. For firms that produce complex assets, the ability to predict with accuracy when maintenance overhauls should occur can provide both an operational and economic competitive advantage. This research evaluates the hypothesis that the accuracy of predictive maintenance models for complex assets can be improved with the addition of historical operational data and failure modes can be more clearly identified by examining primary sensor data. This hypothesis is tested through data analysis on predictive maintenance models used by commercial turbofan jet engines. Because some engines have operated for decades, their entire operational records are not in the appropriate digital format and not utilized by current models. This research identifies alternate, available sources of this data. The additional data sources were processed and incorporated into the existing predictive maintenance models. The addition of the operational data sources did not reduce the error in the model used to forecast the useful life of assets for preventative maintenance, which suggests that the current coverage provided by existing data is sufficient. The examination of primary sensor data isolated one component that displayed age-related degradation and maintenance costs. / by Brian Gabriel Gaudio. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 72-73). / As companies expand and innovate, it is sometimes prudent to be conservative when incorporating new products and technologies. Instead of constructing buildings to store new products, firms may look to optimize re-allocation of space within existing facilities to fit more products and save on one-time capital outlay. In this research, the Distribution Center (which also picks and packs products according to incoming orders) is looking to optimize utilization of offsite storage (and transportation) costs in the face of growing product demand. The DC is situated in a region of High Real Estate cost, and there is potential to increase utilization of offsite leased storage at Low Real Estate cost areas. To investigate potential changes, research will be divided into two parts: Part One looks to optimize product storage within the current Distribution Center and Offsite Warehouse network, by developing a model that incorporates product demand, product sizes and replenishment frequency. Part Two utilizes the built model to investigate alternative offsite solutions, taking into consideration Real Estate costs, transportation frequency and other factors. Previous research papers have looked at the two parts separately, whilst this research aims to link the two parts together. Finally, a simple and easy to use decision-support tool was developed that allows users to periodically review and adjust product allocation based on product information, demand and Real Estate costs. / by Denton He. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 84-86). / Li & Fung works with over 10,000 factories distributed across 50 countries to design, produce, and deliver hard- and soft- goods to over 2,000 apparel and consumer goods customers. An increasingly prevalent focus of the industry, driven both by regulation and consumer preferences, is to measure, benchmark, and reduce the overall environmental impact of the supply chain. Currently the measurement mechanisms in place rely on a traditional two-phase approach involving factory self-reporting and verification via independent audits. The scope of this project is to assess the efficacy of currently available measurement data in order to inform the requirements for real-time collected data. This project will be broken into four phases. First, existing industry data sources will be described and evaluated in order to assess data quality, understand requirements, and provide recommendations for future data collection. Second, the features of the data will be analyzed in order to develop an understanding of the underlining relationships. Third, using a set of selected features from the second phase, a predictive clustering algorithm for factory-level resource efficiency will be developed and used to benchmark factories. Finally, an analysis will be performed to evaluate the requirements of real time data and how real-time data could improve the benchmarking tool and future tools and services. / by Jordan Riley Landis. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 101-104). / Raytheon's Circuit Card Assembly (CCA) factory in Andover, MA is Raytheon's largest factory and the largest Department of Defense (DOD) CCA manufacturer in the world. With over 500 operations, it manufactures over 7000 unique parts with a high degree of complexity and varying levels of demand. Recently, the factory has seen an increase in demand, making the ability to continuously analyze factory capacity and strategically plan for future operations much needed. This study seeks to develop a sustainable strategic capacity optimization model and capacity visualization tool that integrates demand data with historical manufacturing data. Through automated data mining algorithms of factory data sources, capacity utilization and overall equipment effectiveness (OEE) for factory operations are evaluated. Machine learning methods are then assessed to gain an accurate estimate of cycle time (CT) throughout the factory. Finally, a mixed-integer nonlinear program (MINLP) integrates the capacity utilization framework and machine learning predictions to compute the optimal strategic capacity planning decisions. Capacity utilization and OEE models are shown to be able to be generated through automated data mining algorithms. Machine learning models are shown to have a mean average error (MAE) of 1.55 on predictions for new data, which is 76.3% lower than the current CT prediction error. Finally, the MINLP is solved to optimality within a tolerance of 1.00e-04 and generates resource and production decisions that can be acted upon. / by Hans Nowak II. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering

Thesis: M.B.A., Massachusetts Institute of Technology, Sloan School of Management, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 / Thesis: S.M., Massachusetts Institute of Technology, Department of Mechanical Engineering, in conjunction with the Leaders for Global Operations Program at MIT, May, 2020 7102 Sloan School of Management. / Cataloged from the official PDF of thesis. / Includes bibliographical references (pages 103-106). / As automation continues to gain prevalence within the retail industry, informed decision-making by users of robotic systems is critical for management of throughput and operating expenditures. On robotic fulfillment floors, obstructions such as fallen product and deactivated robots can degrade robotic floor throughput by blocking access to product, forcing robots to re-route, and increasing worker idle time. Workers can walk onto the floor to address obstructions during operation, but such entry affects robot movement and can undermine the original intention of restoring throughput. This project aims to provide insight into the cost-benefit tradeoff of resolving obstructions to enable task prioritization and reduce unnecessary floor entry during operation, thereby improving system performance and reducing operating costs. We introduce a novel framework for modeling floor entry to determine the "value" of resolving an obstruction and apply an agile approach to rapidly develop and pilot a software tool for delivery of model recommendations in the field. During the treatment shifts, z-scores of measured pick work unavailability (our chosen performance metric, for which a reduction is indicative of improved throughput), were -0.72, -1.04, and -0.16 as compared with a control sample of similar shifts. The approximate fraction of obstructions resolved during non-operation increased by a factor of three, with recommendation adherence measurements indicating that the increase was driven by elimination of unnecessary (as determined by the model) floor entries during operation. While the sample size was not large enough to achieve a statistically significant outcome, these results offer useful insights regarding future analytical work, testing, and associated organizational changes. / by Ankur Amlani. / M.B.A. / S.M. / M.B.A. Massachusetts Institute of Technology, Sloan School of Management / S.M. Massachusetts Institute of Technology, Department of Mechanical Engineering