2016 INFORMS Annual Meeting Program

SC86

INFORMS Nashville – 2016

3 - Allocation Of Medical Interventions In Outbreak Control: The Case Of Ebola Virus Farbod Farhadi, Roger Williams University, 1 Old Ferry Road, Bristol, RI, 02809, United States, ffarhadi@rwu.edu The outbreak of Ebola in 2014 in western Africa is one of the fastest and deadliest outbreaks in history of viral diseases, causing a reported 28 thousand suspected cases and over 11 thousand deaths, according to WHO, leading to over 70% fatality rate. Further outbreaks of the disease may occur in the future and fast and effective containment strategies to control the spread is vital. In this study a model for efficient allocation of medical interventions for outbreak containment is presented. 5 - Cyclic Physician Scheduling Using Goal Programming Hamoud Sultan Bin Obaid, PhD Student, University of Oklahoma, 1027 E Brooks St., Apt E, Norman, OK, 73071, United States, hsbinobaid@gmail.com, Theodore B Trafalis A two-phase approach to construct a three-month schedule for physicians at an outpatient clinic is proposed. The goal of the proposed model is to minimize the variability of clinic and surgery sessions over the three-month period and utilize resources. From mathematical point of view, the goal is to reduce the complexity of solving this problem. The data used in this problem is obtained from King Khaled Eye Specialist Hospital in Saudi Arabia. Manufacturing III Contributed Session Chair: Zahra Sedighi Maman, Auburn University, Auburn, AL, 36849, United States, zzs0016@auburn.edu 1 - Minimum Cost Allocation Of Quality Improvement Targets: The Effects Of Forgetting And Knowledge Decay Didun Peng, Purdue University, 610 Purdue Mall, West Lafayette, IN, 47906, United States, peng67@purdue.edu Weijia Wang, Robert Plante, Jen Tang This paper incorporates knowledge depreciation in two dimensions of learning: forgetting in autonomous learning and induced learning. We first present a comprehensive quality cost progress function to account for both learning and forgetting effects, where the forgetting effects are imbedded in the progress function components of accumulated production and induced learning. Within the context that a manufacturer allocates quality improvement targets to its suppliers, an optimization model is developed to allocate induced learning activities that minimize the total system cost. A numerical examples of an internal supplier process is used to demonstrate the model. 2 - Sustainability and Changeability In Manufacturing System Shima Ghanei, University of Minnesota -Duluth, 105 Voss Kovach Hall,1305 Ordean Court, Duluth, MN, 55812, United States, ghane009@d.umn.edu, Tarek Al Geddawy Changeable Manufacturing Systems (CMS) are designed to quickly adapt to changing market requirements by transition from a configuration to the next. Not only is the reconfiguration cost dependent on degree of system convertibility and scalability, but also dependent on what time of the year during which it is performed, since energy pricing changes within and between seasons. This paper presents a new linear mixed integer mathematical model to maximize sustainability of CMS on the tactical level. It is solved by CPLEX solver in GAMS software to analyze influence of volatile energy pricing and variable demand on system convertibility and scalability which can affect layout configuration selection. 3 - Printing The Future: Using Analytics To Advance Additive Manufacturing Sarah Powers, Oak Ridge National Laboratory, One Bethel Valley Rd., P.O. Box 2008, Oak Ridge, TN, 37831, United States, powersss@ornl.gov Recent advances in additive manufacturing have led to many success stories of large 3D printed objects and leave the industry poised for rapid growth. This work describes a multi-pronged approach for data discovery, engaging multiple analytic tools as well as a framework to ingest and house the data itself in an effort to identify areas for process improvement and promote the potential for advanced defect detection. SC86 GIbson Board Room-Omni

4 - A Short Note On The Effect Of Sample Size On The Estimation Error In Cp Zahra Sedighi Maman, Auburn University, Auburn, AL, 36849, United States, zzs0016@auburn.edu, William Murphy, Saeed Maghsoodloo, Fatemeh HajiAhmadi, Fadel Megahed Process Capability indices such as Cp are used extensively in manufacturing industries. In practice the parameter for calculating Cp is rarely known and is frequently replaced with estimates from an in-control reference sample. This study explores the optimal sample size required, with some practical tools to achieve a desired error of estimation using absolute percentage error of different Cp estimates.

Sunday, 3:10PM - 4:00PM

Keynote Davidson Ballroom A-MCC Optimizing the Kiel Canal – Online Routing of Bidirectional Traffic Keynote Session Michael Trick, IFORS President, Carnegie Mellon University, Pittsburgh, PA 15213-3890, trick@cmu.edu 1 - Optimizing The Kiel Canal – Online Routing Of Bidirectional Traffic Rolf H Mohring, Beijing Institute of Scientific and Engineering Computing, Beijing, China, rolf.moehring@me.com We introduce, discuss, and solve a hard practical optimization problem that deals with routing bidirectional traffic on the Kiel Canal, which is the world’s busiest artificial waterway with more passages than the Panama and Suez Canal together. The problem arises from scarce resources (sidings) that are the only locations where large ships can pass each other in opposing directions. This requires decisions on who should wait for whom (scheduling), in which siding to wait (packing) and when and how far to steer a ship between sidings (routing), and all this for online arriving ships at both sides of the canal. The lecture is based on joint work with Elisabeth Lübbecke and Marco Lübbecke. Keynote Davidson Ballroom B-MCC Creating the Exascale Ecosystem for Science Invited: Plenary, Keynote Invited Session Bogdan Bichescu, The University of Tennessee, Knoxville, TN 37996-0525, bbichescu@utk.edu 1 - Creating The Exascale Ecosystem For Science Jeff Nichols, Oakridge National Laboratory, Oak Ridge, TN, United States, malonelt@ornl.gov The way we tackle grand challenge science questions at the national scale has changed over the past two decades with the advent of both modeling and simulation (M&S) and “big data” becoming more recognized and supported discovery paradigms. In fact, most large scientific problems today are solved as integrated solutions of experiment, theory, M&S, and data analytics. The past several decades of high performance computing have focused on delivering compute intensive systems and their performance measured by how fast they can accomplish a simple matrix multiply (e.g., high performance linpack). Today’s complex workflows require not only compute intensive capabilities, but also capabilities that target data analytics approaches such as deep learning, graph analytics, or map reduce. In this talk I will describe several scientific areas that require an integrated approach and discuss the ecosystem [ORNL’s Leadership Computing Facility (OLCF) and its Compute and Data Environment for Science (CADES)] that we have created. We continue to invest in the evolution of this ecosystem to enable successful delivery of important scientific solutions across a broad range of disciplines.

94