9:10am-10:00am
Designing the Supply Chain Network to Withstand Market Volatility: A Four-Step Methododology Toby Brzoznowski, Executive Vice President, Llamasoft, Inc.
A slumping economy and volatile commodity prices are two recent factors that have led many companies to realize that their supply chain operations may not be optimized. It is one thing to optimize operations for existing market conditions, but it takes a much deeper analysis to design a supply chain that can withstand fluctuating market conditions. This deeper analysis requires multiple techniques and solution algorithms. In this presentation, Brzoznowski will share a methodology for strategic network planning that has evolved through many client projects. This methodology consists of four incremental steps that help ensure an optimal supply chain network design. The discussion will include case studies to help illustrate approaches and lessons learned, including:
• Capacity and distribution analysis for a leading retailer,
• Production and inventory planning for a global beverage company,
• Transportation planning and asset optimization for a major petroleum producer,
• Service part distribution optimization for a global automotive manufacturer.
10:30am-11:20am
Lessons Learned from the Implementation of a Logistics Deployment Decision-Support System at Pfizer: Preparing for Network Uncertainties Tan Miller, PhD, Harper Professor of Global Supply Chain Management, Director of the Global Supply Chain Management Program, College of Business Administration, Rider University
To manage a large-scale distribution network effectively, a firm must have a decision support system (DSS) that supports strategic, tactical and operational decisions. In this presentation, Miller will describe a DSS implementation at Pfizer that supported daily logistics inventory deployment and shipping decisions across the firm’s entire US consumer healthcare network. Pfizer Logistics developed this DSS in response to a sudden large increase in inventory on the firm’s consumer healthcare network – an unforecast buildup that quickly surpassed the storage capacities of the US distribution network. The rapid development of this DSS allowed Pfizer to maintain high customer service levels while operating with 50% of its inventory in off-site, overflow DC locations. Miller will discuss how its long-standing commitment to DSS systems positioned Pfizer to respond quickly to these short-term crisis conditions. He will highlight this system’s methodologies that have general applicability to other firms, and provide an overview of the complete logistics DSS implemented at Pfizer over approximately ten years.
11:30am-12:20pm
Solving a Strategic Supply Chain Network Design Decision John L. Haller, Jr., Account Manager and Analytic Business Consultant, Strategic Planning and Modeling (SpaM), Hewlett-Packard Company
The Strategic Planning and Modeling (SPaM) group at Hewlett Packard has assisted in the design of many different supply chain networks. Depending on the complexity of the modeling and the number of alternatives considered, SPaM uses different tools to find the recommended solution. In this talk, Haller will focus on the first tool used (and most of the time the only tool needed), a spreadsheet model. The HP team has found this approach superior in addressing the strategic questions that the company considers. Haller will NOT discuss commercial optimization tools; rather he will focus on how SPaM: structures and frames the problem, uses a rough cut analysis to prioritize data collection and model complexity, designs the spreadsheet model that builds insight and answers the specific in-scope decisions, considerers sensitivities that address risk or uncertainty, and develops a communications plan for the recommended the solution. Examples of where this approach has been successful will be given.
2:00pm-2:50pm
Reducing Carbon Emissions in Logistics: A Solution and Lessons Learned from a Pilot Study Kaan Katircioglu, PhD, Research Relationship Manager (Travel & Transportation), Research Staff Member and ODIS Consultant – Supply Chain, Climate Change & Sustainability Management, IBM T. J. Watson Research Center
As a large computer manufacturing company, IBM makes millions of shipments across the world every year. One of the challenges is to accurately measure the carbon emission associated with these shipments, and identify cost-effective ways to reduce it. IBM Research developed a web-based software that can calculate carbon emissions based on shipment records, analyze the trade-offs between carbon and other metrics such as cost and service levels, evaluate the impacts of decisions such as service level changes, shipment consolidation, alternate fuel types or transportation modes, sourcing and packaging choices. The team analyzed IBM’s worldwide shipment data using this solution. Katircioglu will share findings from this study; he will describe the most effective levers for carbon reduction that were found, and present win-win scenarios that can reduce both carbon emissions and transportation costs. He will also articulate the lessons learned and make recommendations to companies that want to take similar initiatives.
3:30pm-4:20pm
Green Forum at Eroski: Optimizing Logistics Activities with Environmental Criteria Javier Faulin, Associate Professor of Statistics and OR, Public University of Navarre, Spain
In today’s highly competitive business arena, green logistics issues are gaining interest. This talk will analyze how some logistics managers who established a Green Forum in a Spanish retail company, called Eroski, could lead some initiatives in sustainable logistics by incorporating environmental management principles into their daily decision-making process. Some of the problems tackled in the forum were:
• Discussion of an optimization vehicle routing problem with a goal of minimizing environmental impact of the whole distribution process.
• Statistical study of the real efficiency of formulating special additives to the fuel used by the Eroski distribution fleet.Those additives purported to reduce the polluting emissions of Eroski trucks.
• Development of a general study about the location of future Eroski shops in cities, towns and villages according to some sustainable criteria of minimizing noise impact, pollution and neighbors’ complains.
• Optimum design of protocols of loading and unloading vehicles in order to speed those logistics processes considering efficiency and sustainability criteria.
Track 11, Supply Chain Management - Tuesday 9:10am-10:00am
Implementing Supply Routing Optimization at Dell Jérémie Gallien, PhD, Associate Professor of Operations Management, MIT Sloan School of Management
For over a year, Dell has been using a new optimization-based control system to generate supply routing decisions for its US-bound monitors. The system was developed by a partnership between Dell’s Supply Chain Command Center and a team of university researchers, and is the focus of this talk. Gallien will discuss the development, implementation and impact of this new supply routing process. In the first phase of the partnership, Dell created a new job definition focused solely on supply routing and implemented a supporting visualization tool. In a second phase, a decision support system relying on an MIP formulation was implemented, overcoming two main technical challenges: (i) the estimation of expected shortages as a function of the various routing decisions considered, accounting for actual forecast quality; and (ii) the estimation of a meaningful shortage cost. This new methodology is estimated to have reduced Dell’s inventory re-positioning costs for monitors by about 60%. Key take-aways for the audience are:
• Description of the inner workings of a large supply-chain operated by a firm recognized for being both a leader in the computer industry and a pioneer in supply-chain management;
• Description and discussion of an OR model successfully implemented in this context, which has resulted in a significant and measurable reduction of Dell’s transportation costs;
• Discussion of the main technical and organizational challenges encountered during that implementation, and how they were overcome.
10:30am-11:20am
Reducing Costs to Compete through Drayage Optimization Xiaoqing (Peter) Sun, PhD, Senior Logistics Engineer, Schneider National, Inc.
Intermodal is a major part of Schneider National business offerings. Efficient drayage operations can reduce empty miles, better meet service requirements and protect the environment. Dray dispatching is a complicated process that has to meet operations requirements such as appointment windows, rail cutoffs, transit times, driver hours, hazmat and time at home. Drayage optimization is part of the effort aimed at efficiently dispatching drayage operations . In this talk, Sun will present:
• Brief description of the mathematical algorithm;
• Data structure that allowed parallel processing;
• Business impact including loaded ratio, driver utilization, and service quality;
• Change management in deploying the tool;
• Practical challenges such as data inaccuracies and uncertainties;
• Future projects.
11:30am-12:20pm
Clean Sheet Airline Scheduling Alex Heinold, MS, Manager of Schedule Development, Southwest Airlines
Southwest Airlines is well known for its simplified approach to flying. In keeping to this approach, Southwest has avoided complexity-driving strategies like large connecting hubs and flying multiple types of aircraft, strategies that normally consume much of a typical airline’s OR attention. While these high-level strategies do simplify many aspects of the airline, complexity still shows up in different ways and opportunities for OR work abound. This presentation will follow a nine-year journey of OR work at Southwest in the area of flight scheduling. The presentation will cover:
• How Southwest’s unique network design and scheduling philosophy led to the development of a constraint programming based approach to “clean-sheet” scheduling.
• Insights on the long process of calibrating the model and cultural issues that arose from the new business practices that were enabled by the new tools.
• The transition from the original constraint programming model to the current linear programming approach.
• The ways that this OR work has contributed to Southwest’s success, including a more attractive schedule, increased asset utilization and improved ability to react quickly to market changes.
3:10pm-4:00pm
Delivering OR-Based Freight Optimization in the On-Demand World Christopher M. Johnson, BS, Freight Optimization Services R&D Manager, LeanLogistics, Inc.
The on-demand (SaaS) model has established itself as a preferable method for providing software. The delivery of technology as a service offers unique opportunities for the development and execution of solutions. As with most transportation management systems, freight optimization is a key component, and delivering valuable, OR-based solutions continues to be critical as companies streamline the transportation portion of their supply chain. Key topics in this talk include:
• An overview of current, constraint-based freight optimization capabilities;
• A brief demonstration of solving real-world freight optimization problems;
• A discussion of the specialized technology and infrastructure used to deliver resource-intensive optimization within the Software as a Service (SaaS) model;
• A case study detailing the usage of the freight consolidation engine.
4:10pm-5:00pm
Practical Considerations in Optimizing MRO Parts Inventory at the Los Angeles Metropolitan Transit Authority: A Case Study Arnold Greenland, PhD, IBM Distinguished Engineer, IBM Global Business Services
Optimization of maintenance, repair and overhaul (MRO) parts inventory has been studied extensively because of its centrality to many businesses and because of the challenges of practical implementation. This talk will focus on how accurate modeling of the complex variability in demand and supply can improve inventory policies. Greenland will reference a recent pilot project done for the Los Angeles Metropolitan Authority (Metro) to optimize MRO parts inventory. He will compare inventory planning results using the existing LA Metro methods, a set of leading practices used in the pilot, and a set of proposed alternative methods relying on manufacturer supplied failure/risk data. Finally, the presentation will cover the inventory management issues encountered in this project such as designing and implementing the right organization structure, business processes, IT capabilities and performance metrics.
Supply Chain Management Presentations in Other Tracks
Track 7, Optimization - Monday 9:10am-10:00am
Approximate Dynamic Programming for Spare Parts Management Hugo P. Simão, PhD, Senior Operations Research Engineer, CASTLE Laboratory, Department of Operations Research and Financial Engineering, Princeton University
Your company stores and distributes spare parts directly to customers. Some parts are affordable and have high demand, but others are expensive and have extremely low demand. Your customers can solicit parts anywhere in the world and in some cases they have to be delivered in less than six hours. Demand will be increasing for the next several years. The distribution network comprises of a few major warehouses and several regional warehouses and there is a complex third-party transportation network serving the warehouses and the customer locations. Your challenge is to manage the parts inventory at all the warehouses so that a minimum level of customer service is met, without exceeding your budget constraints. Simão will show how approximate dynamic programming can provide a decision-making tool that can help you meet that challenge. The approach will be illustrated through the example of an actual aircraft spare parts management system.
2:00pm-2:50pm
A Decomposition Approximation Approach for Multi-Echelon Service Level Analysis at Hewlett-Packard Divya Mangotra, PhD, Analytical Business Consultant, Hewlett-Packard Company
The Strategic Planning and Modeling group at HP led an investigation to evaluate whether the company is holding the right amount of inventory for one of its key businesses. This required modeling a multi-echelon supply chain to understand how inventory held at upstream DC impacts the service level observed at its end customers. Divya will discuss the problem structure; limitations of a commercial inventory optimization software to model it and a decomposition approximation approach that is developed to solve the problem. She will describe in detail the modeling approach for estimating the additional wait time (delay in order fulfillment experienced by channel when there is a stock-out at the DC) and conclude the discussion by a brief mention of the advantages and disadvantages of the decomposition approach.
Track 14, Methodology Tutorials – Tuesday 9:10am-10:00am
2008 Daniel H. Wagner Prize Winner: Managing Inventory in Supply Chains with Nonstationary Demand John J. Neale, PhD, Assistant Professor, Operations & Technology Management, Boston University; Sean Willems, PhD, Founder and Chief Scientist, Optiant, Inc. and Associate Professor, Boston University
Increasingly, companies are experiencing demand for their products that is not only uncertain but also nonstationary. This means the demand is both unpredictable and changing over time. Examples include short lifecycle products at Hewlett-Packard, seasonal products at Kraft Foods and “hockey stick” demand patterns at Dell. In this talk, Neale and Willems will discuss a practical approach for managing inventory in the face of nonstationary demand. They will present a model that can be used to address the strategic question of where to locate inventories in the supply chain and the tactical question of how to change inventory targets over time. They will also describe two successful applications of this model. The first is a tactical implementation to support monthly safety stock planning at Microsoft. The second is a strategic project to evaluate the benefits of an inventory pool at Case New Holland.
