Stock Management Complexities: Model Formulations and Human Experiments

Hakan Yaşarcan
Accelerated Learning Laboratory, Australian School of Business (Incorporating AGSM), University of New South Wales, Sydney, NSW 2052, Australia

Abstract: This is a two-part talk based on two related research work: Delays involved in stock management systems can create unwanted oscillatory behaviour if they are not properly accounted for in the decision heuristics. There are many studies on control formation delays. However, there are no explicit studies on delays involved in measuring/perceiving stock levels. Thus in the first part, we discuss the difficulty caused by the presence of measurement delays and suggest a decision formulation that handles it. In the second study, we show that dynamic complexities involved in a relatively simple stock management task, a basic inventory management simulation game, create learning difficulties for trainees. We propose gradual-increase-in-complexity approach to overcome these difficulties and improve the effectiveness of the simulation based training and the learning curve.

Bio:
Hakan Yasarcan (BS, MS, PhD) earned all his degrees in Industrial Engineering. He received his PhD from Bogazici University, Istanbul, in 2003. He initially became interested in System Dynamics after meeting Prof. Yaman Barlas and joining the organising committee for the Fifteenth International System Dynamics Conference of the System Dynamics Society (Istanbul, 1997). He then became a member of the Socio-Economic System Dynamics Laboratory (SESDYN) at Bogazici University and assisted Prof. Barlas in the management of SESDYN for six years. Dr Yasarcan has many years of experience as research and teaching assistant and as a faculty member. He was previously working as an Assistant Professor of Industrial Engineering at Eastern Mediterranean University in Northern Cyprus until August 2006. He then relocated to Australia to join the Australian School of Business as a post doctoral research fellow at the Accelerated Learning Laboratory (ALL) directed by Prof. Robert Wood. He has a number of publications on generalised stock control formulations and dynamic goal management. His current research focuses on interactive simulation games and dynamic decision making. He has also been teaching and practicing Sahaja Yoga for the last seven years.

Cooperative Advertising and Pricing in a Dynamic Stochastic Supply Chain: Feedback Stackelberg Strategies

Suresh P. Sethi
University of Texas at Dallas

Abstract: Cooperative (co-op) advertising is an important instrument for aligning manufacturer and retailer decisions in supply chains. In this, the manufacturer announces a co-op advertising policy, i.e., a participation rate that specifies the percentage of the retailer's advertising expenditure that it will provide. In addition, it also announces the wholesale price. In response, the retailer chooses its optimal advertising and pricing policies. We model this supply chain problem as a stochastic Stackelberg differential game whose dynamics follows Sethi's stochastic sales-advertising model. We obtain the condition when offering co-op advertising is optimal. We provide in feedback form the optimal advertising and pricing policies for the manufacturer and the retailer. We contrast the results with the advertising and price decisions of the vertically integrated channel, and suggest a method for coordinating the channel.

Bio:
Suresh P. Sethi is Charles & Nancy Davidson Distinguished Professor of Operations Management and Director of the Center for Intelligent Supply Networks at The University of Texas at Dallas. He has written 5 books and published more than 300 research papers in the fields of manufacturing and operations management, finance and economics, marketing, and optimization theory. He teaches a course on optimal control theory/applications and organizes a seminar series on operations research topics. He serves on the editorial boards of several journals including Production and Operations Management, SIAM Journal on Control and Optimization, and Automatica. Recent honors include: POMS Fellow (2005), INFORMS Fellow (2003), AAAS Fellow (2003), IEEE Fellow (2001), IITB Distinguished Alum (2008). Two conferences were organized and two books edited in his honor in 2005-6.

Optimal Search for a Moving Target with the Option to Wait

Emin Karagözoğlu
Maastricht University, Department of Economics, The Netherlands

Abstract: We investigate the problem in which an agent has to find an object that moves between two locations according to a discrete Markov process (see Pollock, 1970). At every period, the agent has three options: searching left, searching right, and waiting. We assume that waiting is costless whereas searching is costly. Waiting can be useful because it could induce a more favorable probability distribution over the two locations next period. We find an essentially unique (nearly) optimal strategy, and prove that it is characterized by two thresholds (as conjectured by Weber, 1986). We show, moreover, that it can never be optimal to search the location with the lower probability of containing the object. The latter result is far from obvious and is in clear contrast with the example in Ross (1983) for the model without waiting. We also analyze the case of multiple agents. This makes the problem a more strategic one, since now the agents not only compete against time but also against each other in finding the object. We find different kinds of subgame perfect equilibria, possibly containing strategies that are not optimal in the one-agent case. We compare the various equilibria in terms of cost-effectiveness.

Bio:
Emin Karagözoğlu received BA and MA degrees in economics from Boğazici University in 2001 and 2003. Between 2000 and 2003, he was a research assistant in the Center for Economics and Econometrics and teaching assistant in the Department of Economics. Afterwards, he joined The Pennsylvania State University (USA), where he completed Ph.D. coursework and received another MA degree in economics. Since December 2006, he has been working on his Ph.D. thesis in Maastricht University (The Netherlands). His thesis aims to provide different approaches (e.g., noncooperative, cooperative, experimental and normative) to bankruptcy problems. His research interests cover noncooperative game theory, distributive justice & fair division, conflict resolution, operations research & applied probability and cognitive economics.

On the Online Track Assignment Problem

Marc Demange
ESSEC Business School, Cergy Pontoise, France

Abstract: This work deals with the computational complexity of some online shunting problems. We analyze the following problem. Consider a train station consisting of a set of parallel tracks. Each track can be approached from one side only or from both sides and the number of trains per track may be limited or not. The departure times of the trains are fixed according to a given timetable. The problem is to assign a track to each train as soon as it arrives to the station so that it can leave the depot on time without being blocked by any other train. We show that this problem can be modeled with online coloring of graphs. Depending on the constraints, the graphs can be overlap graphs (also known as circle graphs) or permutation graphs, and the coloring can be bounded or classical.

Bio:
Marc Demange received his HDR from the department of Computer Science at Paris-Dauphine University and Ph.D degree again in Computer Science from Paris I University. Then he did a research master (DEA) in Modeling and Mathematical Methods for Economy at Paris I University-Ecole Polytechnique. He also did Agrégation (French diploma for teaching) in probability. Marc Demange is now professor of Operational Research at ESSEC Business School - Paris - Singapore. He is also the dean of research of the school. His work concerns combinatorial optimization and in particular complexity theory, polynomial time approximation, online coloring and inverse combinatorial optimization. He has published more than 30 papers in refereed journals in these areas.

Evaluation Overreaction to Backlog as a Behavioral Cause of the Bullwhip Effect

Rogelio Oliva
Texas A&M University
USA

Abstract: We evaluate, in an experiment with the Beer Distribution Game, a complementary behavioral source of the bullwhip effect that has been previously ignored in the literature: overreaction to backlogs. By separating the estimation of the response to inventory and backlog, we find that players treat backlog differently than inventory. Contrary to our expectations, players do not over-order when in backlog; instead, they have a measured response, saturating order adjustment and limiting the amount of amplification they introduce in the order stream. We also find stronger evidence than previous studies that players underestimate the supply line, leading to a more unstable ordering policy. Using a simulated order stream, we find that players display bounded rationality and that their estimated decision policy is not different in form and performance than a policy that, with the information cues available in the Beer Distribution Game (inventory position and orders), minimizes local cost. Players' estimated ordering policy, however, aims to maintain higher inventory levels, leading to increased order amplification and costs for upstream echelons. Hence, the estimated ordering policy presents strong behavioral components: it ignores the supply line and under-reacts to backlog while aiming for higher than necessary inventory levels. We conclude by discussing the implications of these findings for future research and practice in supply chain management.

Bio:
Rogelio Oliva is Associate Professor of Information and Operations Management at the Mays Business School at Texas A&M University. His current research interests include service operations, behavioral decision making in supply chains, and the transition that product manufacturers are making to become service providers. Dr. Oliva has published in Management Science, California Management Review, European Journal of Operational Research, System Dynamics Review, and the Int. Journal of Service Industry Management. A native of Cd. Valles, Mexico, he holds a B.E. in Industrial and Systems Engineering from the Monterrey Technological Institute (ITESM) in Mexico, an M.A. in Systems in Management from Lancaster University (UK), and a Ph.D. in Operations Management and System Dynamics from the Massachusetts Institute of Technology. Prior to joining the Mays faculty, Professor Oliva taught for six years at the Harvard Business School and for three years at ITESM in Mexico.

Survivable Network Design Problems and Polyhedra

Ali Ridha Mahjoub
LAMSADE, Université Paris-Dauphine
France

Abstract: The introduction of fiber optic technology in telecommunication has increased the need of designing survivable networks. Survivable networks must satisfy some connectivity requirements. That is networks, which are still functional after the failure of certain links. More precisely, we are given a graph G = (V, E) with costs on the edges. For each node v there is a nonnegative integer r(v), called connectivity type of v, that represents the importance of communication from and to node v. The survivable conditions require that between every pair of nodes (s, t) there are at least min{r(s), r(t)} edge-disjoint paths. The survivable network design problem is to determine a subgraph of G that minimizes the total cost subject to the survivable conditions. In this talk, we will discuss some variants of this problem. First we consider the 2-edge connected subgraph problem, the case where r(v)=2 for every node. We characterize the graphs for which the linear relaxation of the problem is integral, and present some algorithmic consequences of that characterization. In particular, we show that the so-called F-partition inequalities can, in some cases, be separated in polynomial time. Then we consider the problem when r(v)=1 or 2 for every node v. This case is of particular interest to the telecommunication industry. We present a class of inequalities, also called partition inequalities, valid for the problem in this case. And we show that the separation problem for these inequalities reduces to the minimization of a submodular function, and can then be solved in polynomial time. We finally discuss some computational issues of these inequalities and present some extensions when length constraints are considered in the network.

Collection System Design Problem for Product Recovery

Necati Aras
Boğaziçi University,
Istanbul, Turkey

Abstract: In this talk, I will discuss about variants of the collection system design problem for product recovery. Product recovery has emerged in the last decade as a popular business strategy within the context of sustainable development. The type of the product recovery is dependent on the condition of a return. The possibilities are repairing, refurbishing, remanufacturing, cannibalization, and recycling. In all the variants of the problem, the company involved in product recovery collect used products called cores from product holders under two different policies: pick-up policy and drop-off policy. Each product holder has an inherent willingness to return, and makes the decision on the basis of the financial incentive offered by the company under the pick-up policy and both the incentive and the distance traveled to the collection center under drop-off policy. The incentive depends on the condition of the returned item referred to as return type. We formulate a mixed-integer nonlinear facility location-allocation model to find both the optimal locations of a predetermined number of collection centers, the optimal incentive values for different return types, and the number of vehicles and their load compositions. Since the problem is NP-hard, we propose a heuristic method to solve medium and large size instances. Then we consider an extension of this problem where the company is subsidized by the government for each core collected. We formulate two bilevel programming (BP) models describing the subsidization agreement between the government and the company. These are supportive and legislative BP models. In both models, the outer problem of the BP formulation involves the government, which is the leader and wants to minimize the unit subsidy. The company is the follower in the inner problem, and tries to maximize its net profit from the cores subject to the government’s subsidy decision. In the supportive model, the company itself is not bound by the minimum collection rate targeted by the government, thus the amount of collected cores may not be sufficient. The government tries to resolve this situation with increased subsidization. The legislative model assigns the minimum collection rate responsibility to the company, but also entitles it to a certain profitability ratio guaranteed by the government. The solution methodology proposed for both models consists of Brent's method for the outer problem and a tabu search heuristic solving the inner problem. Its effectiveness is tested in computational experiments.

Bio:
Necati Aras is an assistant professor in the Department of Industrial Engineering at Boğaziçi University.

Collection System Design Problem for Product Recovery

Necati Aras
Boğaziçi University,
Istanbul, Turkey

Bio:
Necati Aras is an assistant professor in the Department of Industrial Engineering at Boğaziçi University.

Inventory, Scheduling and Lead Time Decisions in Supply Chains

Onur Kaya
Koç University,
Istanbul, Turkey

Abstract: In our study, we consider a multi-item manufacturer and its suppliers with different supply chain architectures in a stochastic environment. We consider the scheduling, lead-time quotation and inventory problems in this model. The manufacturer and the suppliers have to decide which items to produce to stock and which ones to produce to order, how much inventory to keep for make-to-stock items, how to schedule the production of different products and how to quote short and reliable due-dates to arriving customers. In our study, we first focus on a 2-party supply chain and then extend our analysis to more complex settings. We consider several variations of this problem, and design effective heuristics for scheduling and lead time quotation as well as the make-to-order/make-to-stock decision, and to find the appropriate inventory levels for make-to-stock items. We perform extensive computational testing to assess the effectiveness of our algorithms, and to compare the centralized and decentralized models in order to quantify the value of centralized control and information in this supply chain setting.

Bio:
Onur Kaya is an assistant professor in the Department of Industrial Engineering at Koç University. After graduating from the Industrial Engineering Department of METU in 2002, he received his M.S. in 2003 and Ph.D. in 2006 from University of California, Berkeley, Department of Industrial Engineering and Operations Research. He also has an M.A. degree in Statistics from the same university. Onur Kaya’s research interests include stochastic models of production and inventory systems, management of supply chain systems, gaming and contracts in supply chains, dynamic programming, queueing theory, and scheduling.