IE DEPARTMENTAL SEMINARS - ABSTRACTS 2009

October 16, 2009 Friday

Abstract: In this talk, a new variant of the Traveling Salesman Problem with Pickup and Delivery (TSPPD) is presented, in which we consider the handling cost incurred at the customer locations to rearrange the load, as well as the routing cost. The simultaneous optimization of routing and handling costs is difficult and the resulting loading patterns are hard to implement in practice. However, this option makes economical sense in contexts where the routing cost dominates the handling cost. We have proposed some simplified solution methods applicable for such contexts. The first is a two-phase heuristic, in which the tour having minimum routing cost is initially determined by optimally solving a TSPPD, and the optimal handling policy is then determined for that tour, through a dynamic programming formulation. In addition, branch-and-cut algorithms based on integer linear programming formulations are proposed, in which routing and handling decisions are simultaneously optimized, but the handling decisions are restricted to three simplified policies. The formulations are strengthened by means of problem specific valid inequalities. The proposed methods have been extensively tested on instances involving up to 25 customers and hundreds of items. Our results demonstrate the impact of the handling aspect on the customer sequencing and indicate that the simplified handling policies favorably compare with the optimal handling decisions.

Bio:
Maria Battarra completed a 5 years joint BS and MS degree program in Industrial Engineering at the University of Bologna, in 2005. She worked on her thesis, “Heuristic algorithms for Vehicle Routing Problems”, at the University of Maryland, under the supervision of Proff. B.L. Golden, D.Vigo and P.Toth. She is currently a PhD student in Operations Research at the University of Bologna and her PhD is expected to end in Spring 2010. She has been a visiting student at CIRRELT-Canada, where she worked with Proff. J.-F.Cordeau and G.Laporte, and she is currently visiting Boğaziçi University, under the supervision of Proff. K.Altınel and T.Öncan. She is the author of articles published on international books and journals in the field of Operations Research. She works mainly on both heuristic and exact algorithms for variants of the Vehicle Routing Problem.

September 11, 2009 Friday

Abstract: Many decision problems that arise in the context of supply chain management can be formulated as optimization problems. These optimization models typically have parameters, e.g. arrival rates, demand, etc., that are either measured or estimated from a finite amount of data; consequently, parameter estimates always have errors. Optimization models are especially susceptible to these errors since, in trying to maximally exploit the constraints, the optimal solutions typically amplify the errors several fold. The performance of resulting "optimal" decision is, therefore, often poor and unreliable. Robust optimization has recently emerged as a particularly useful methodology for optimizing performance in the presence of data errors. In this talk I will begin with illustrating the basic techniques underlying robust optimization. I will then present robust optimization based models for inventory management, contract selection and call-center staffing and routing.

Bio:
Garud Iyengar received a B. Tech. in Electrical Engg. from IIT Kanpur in 1993 and a Ph.D. in Electrical Engineering from Stanford University in 1998. Since then he has been with the Department of Industrial Engineering and Operation Research Department at Columbia University where he is currently an Associate Professor. He is the recepient of the President's Gold Medal from IIT Kanpur and the NSF CAREER award. His research interests span a number of different topics such as robust optimization and its applications, approximation methods for stochastic systems, approximation methods for auctions and other mechanism design problems, and applications of semidefinite programming.

June 12, 2009 Friday

Abstract: Shift scheduling deals with the assignment of periods of work (shifts) to person in a way that meets organizational as well as legal, health, and social requirements. Shift scheduling is an important task in heavy industries (e.g., steel, chemical) but even more in service Industries (e.g., transport, police, health). Shift scheduling involves several computationally hard scheduling problems. Typically, efficient algorithms exploit some specifics of the situation at hand to run in reasonable time. Those specifics however vary substantially between industries and organizations (e.g. break rules, labour rules). Correspondingly, the adaptation of scheduling algorithms to those specifics becomes a complex problem in itself. In my talk, I will give a brief history on two optimization systems we developed (for cyclic scheduling and for proper design of shifts) and experiences made in adapting them to different specifics. Then I will outline our current work that tries to combine the strength of general optimization engines with the need for easy adaptation to organizational specifics.

Bio:
Assoc. Prof. Dipl. Ing., Dr. Techn., finished his Diploma thesis 1986 at the University of Linz, his PhD 1992 and his habilitation in 2001 at the Vienna University of Technology. He is an expert in shift scheduling since 1990 with a large number of publications in the field. Since 2005, he is member of the board of the Working Time Society. He is one of the founders and the CEO of XIMES 1997, a certified management consultant. He wrote two books in the field of working time and one book on Project-Design.

May 8, 2009 Friday

Abstract: We consider a problem dealing with the efficient delivery of Intensity Modulated Radiation Therapy (IMRT) to individual patients. IMRT treatment planning is usually performed in three phases. The first phase determines a set of beam angles through which radiation is delivered, followed by a second phase that determines an optimal radiation intensity profile (or fluence map). This intensity profile is selected to ensure that certain targets receive a required amount of dose while functional organs are spared. In order to deliver these intensity profiles to the patient, a third phase must decompose them into a collection of apertures and corresponding intensities. In this paper, we investigate this last problem. Formally, an intensity profile is represented as a nonnegative integer matrix; an aperture is represented as a binary matrix whose ones appear consecutively in each row. A feasible decomposition is one in which the original desired intensity profile is equal to the sum of a number of feasible binary matrices multiplied by corresponding intensity values. In order to most efficiently treat a patient, we wish to minimize a measure of total treatment time, which is given as a weighted sum of the number of apertures and the sum of the aperture intensities used in the decomposition. We develop the first exact algorithm capable of solving real-world problem instances to optimality within practicable computational limits, using a combination of integer programming decomposition and combinatorial search techniques. We demonstrate the efficacy of our approach on a set of test instances derived from actual clinical data and randomly generated instances.

Bio:
Z. Caner Taşkın is a PhD candidate at Department of Industrial & Systems Engineering, University of Florida. He got his BS and MS degrees in Industrial Engineering from Boğazici University, İstanbul. Before starting his doctorate study, he worked for ICRON Technologies as a product consultant, where he took role in advanced planning and scheduling (APS) projects for customers in several industries including steel, automotive, electronics and glass manufacturing industries. His current research interests include large-scale optimization, mixed-integer programming, decision making under uncertainty, and medical applications of operations research.

April 21, 2009 Tuesday

Abstract: Based on the recent incidents of avian flu (H5N1) in Asia and the influenza pandemic cases in history (1918, 1957 and 1968) experts believe that a future influenza pandemic is inevitable and likely imminent. Evidence suggests that an efficient and rapid response will be crucial for mitigating morbidity, mortality, and costs to society. Hence, preparing for a potential influenza pandemic has received high priority from governments at all levels (local, state, federal), non-governmental organizations (NGOs), and companies. In collaboration with the American Red Cross, we study the logistics side of the problem, specifically, food distribution logistics during an influenza pandemic. We develop a disease spread model to estimate the spread pattern of the disease geographically and over time, integrate it to a facility location and resource allocation network model for food distribution, and develop heuristics to find near-optimal solutions for large instances of the embedded facility location problem. We run our integrated disease spread and facility location model for the state of Georgia and present the estimated number of infections and the number of meals needed in each census tract for a one year period along with a design of the supply chain network. We analyze the impact of two intervention strategies, namely, school closure and voluntary quarantine; our results indicate that voluntary quarantine may be a better alternative due being more effective and less disruptive. Moreover, we investigate the impact of voluntary quarantine on the food requirement and the food distribution network, and show that its effect on the food distribution supply chain can be significant.

Bio:
Ali Ekici is a Ph.D. candidate in the H. Milton Stewart School of Industrial and Systems Engineering at the Georgia Institute of Technology. He received Bachelor of Science degrees in Industrial Engineering and Mathematics from Middle East Technical University in 2003 and a Master of Science degree in Operations Research from Georgia Institute of Technology in 2006. He was awarded the Best Graduate Research Paper Award sponsored by the Institute of Industrial Engineers Society for Health Systems in 2009 for his paper entitled “Modeling Influenza Pandemic, Intervention Strategies and Food Distribution”. His research experience and interests are in the field of emergency response logistics, network design/expansion, dynamic routing, decentralized decision making and industry applications of scheduling/packing.

April 17, 2009 Friday

Abstract: Hazardous materials (hazmats) are an integral part of our industrialized lifestyle and large quantities of them need to be transported between the points of their production to the points of their consumption or disposal. Although the transportation sector is mostly deregulated, hazmat shipments remain regulated by government agencies in many countries, mainly due to the associated public and environmental risks. One of the policy tools that are available to a government agency is to close certain road segments to the shipments that involve certain types of hazmats. The presentation will focus on three alternative approaches for modeling and solving the regulators’ problem of identifying the most appropriate road segments to apply such curfews while maintaining the economic viability of the sector and minimizing the population exposure to hazmat transport risks. These approaches involve the use of (i) a bilevel programming model, (ii) a path-based formulation, and (iii) a tariff-based model for hazmat network design. Our findings through the implementation of these alternative approaches on the highway networks of the provinces of Quebec and Ontario in Canada will also be discussed.

Bio:
Vedat Verter is a Professor of Operations Management at Desautels Faculty of Management, McGill University. He is also an Associate Member of McGill’s School of Environment and an Adjunct Professor at Telfer School of Management, University of Ottawa. Professor Verter’s research focuses on supply chain design, hazardous materials logistics, sustainable supply chains and healthcare operations management. His work in these four areas is well recognized through top tier journal publications as well as invited presentations around the globe. Professor Verter is the Founding President of the POMS College of Healthcare Operations Management.

March 27, 2009 Friday

Abstract: Manufacturers have increasingly been selling through direct online channels (i.e. selling directly to consumers through their own web sites) alongside traditional retail channels. In this talk, we address a manufacturer's problem of managing dual (direct and retail) sales channels when the channels compete in service. We develop a game-theoretic model of the strategic interaction between the manufacturer and the retailer. We incorporate a detailed consumer choice model in which the consumers consider the delivery lead time in the direct channel and the product availability level in the retail channel in their channel choice. We identify optimal dual channel strategies for the manufacturer, and analyze how these strategies change with respect to changes in the environment. We also present the results of controlled experiments with human decision makers, which we conduct to investigate whether our analytical model makes reasonable predictions of human behavior.

Bio:
Murat Kaya has joined Sabanci University as an assistant professor in Manufacturing Systems & Industrial Engineering Program, in 2007. He received his B.S. degree from Middle East Technical University in 2001, and his M.S. and Ph.D. degrees in Management Science and Engineering from Stanford University in 2002 and 2007, respectively. During his Ph.D. study, Dr. Kaya worked for several projects in Hewlett Packard Research Laboratories in California, USA. Dr. Kaya’s research focuses on game-theoretic modeling in supply chain management. He is also interested in channel management and behavioral economics applications in operations management.

January 23, 2009 Friday

Abstract: The preparation of IV medication at a hospital's inpatient pharmacy is often achieved through the daily processing of few, infrequent batches, each consisting of a large number of individual medication orders. These batches have extended preparation times (cycle-times) due to significant processing, delivery, and queuing time; medication may be prepared for patients up to nineteen hours in advance of administration. These extended cycle-times also increase the probability of wasting medication as patients' treatment may be modified after their medication has been prepared and delivered, but before it has been administered. The staffing schedule must also accommodate the peak workloads caused by the release of large batches of orders to the pharmacy. It was hypothesized that level-loading (i.e., the spreading the scheduled demand for IVs evenly throughout the day) to smooth the arrival of IV medication orders to the inpatient pharmacy would both reduce the mean (average) cycle-time of medication orders (thereby reducing the probability that they are wasted) and permit a more flexible staffing schedule. The performance of current and new operational policies based on this hypothesis were investigated via simulation. Multiple policy improvements were proposed and simulated, ranging from basic and easily applied to the more advanced yet less easily implemented and maintained. Improvements over the current policy were obtained by all new operational policies analyzed. In the best-performing policy, IV medication demand, via the staggering of administration times, is smoothed such that a single resource (compared to the 2 resources currently working simultaneously) is able to prepare all medication orders while dramatically reducing IV medication cycle-time with the significant potential to reduce the amount of wasted medication.

Bio:
Paul Intrevado is a doctoral student in the Desautels Faculty of Management (Operations Management) at McGill University. He previously completed Baccalaureate degrees in Commerce (Management Science concentration) at McGill University and Industrial Engineering at Purdue University and a Master's Degree in Industrial Engineering at Purdue University. Paul's graduate research focuses on the application of systems engineering concepts to healthcare settings in an effort to systematize and improve the efficiency, safety, and quality of healthcare service delivery.

January 09, 2009 Friday

Abstract: Until recently, the logistics literature has addressed the major supply chain drivers, e.g., strategic facility location decisions, tactical inventory management decisions, and operational routing /scheduling decisions, independently due to the nature of their varying time horizons. Hence, most supply chain network design models have overlooked the relationships between the strategic, tactical, and operational elements of a supply chain, resulting in a degree of sub-optimization. For example, most network design models assume direct shipments from a facility to a retailer, resulting in the overestimation of transportation costs; in practice, shipments from a facility to a designated set of retailers is most likely a 'milk-route'. This overestimation of transportation costs results in solutions that open unnecessarily large number of facilities, thereby incurring high fixed location costs. Traditional location models have incorporated routing decisions, which are referred to as location-routing models; location-inventory models have not yet incorporated these routing decisions. By solving the location-inventory-routing model, we capture the complex interactions between the major supply chain drivers and provide the means to estimate the degree of suboptimality imposed on us by the traditional approaches. We present the benefits of integrated supply chain design through extensive computational experiments.

Bio:
Leyla Ozsen is an Assistant Professor in College of Business at San Francisco State University. Prior to SFSU, she was an assistant professor in the Industrial Engineering department at Purdue University. Dr. Ozsen received her Ph.D. in Industrial Engineering and Management Sciences from Northwestern University in 2004, her MS in Management Science and Engineering from Stanford University in 1999, and her BS in Operations Research and Industrial Engineering from Cornell University in 1997. Her research interests include logistics, production planning, and supply chain management.