Informs Annual Meeting 2017

WB29

INFORMS Houston – 2017

5 - Horizontal vs. Vertical Cooperation in Supply Chains under Competition Shuxiao Sun, Doctoral Candidate, Peking University, No.5 Yiheyuan Road Haidian District, Beijing, Beijing, 100871, China, sunshuxiao@pku.edu.cn, Xiaona Zheng We study the impact of horizontal and vertical cooperation with different degree on firms’ profits in two-tier supply chains. The cooperating firm maximizes its own profit and a proportion of its partner’s. We find that, when retailers are asymmetric in variable cost or market base, under horizontal cooperation, the “strong” retailer is always better off while the “weak” one is likely to get worse; under vertical cooperation, the manufacturer prefers cooperation with the “weak” player whereas the supply chain favors “an alliance between giants.” Finally, we extent our model to one manufacturer and multiple retailers. 350E Manufacturing Contributed Session Chair: Sanam Azadiamin, Ohio University, Athens, OH, United States, sa349416@ohio.edu 1 - A MILP Formulation for Combined Planning and Scheduling of Multistage Multipurpose Batch Plants Zhaohui Li, Shanghai Jiao Tong University, 1954 Huashan Road, Shanghai, 20030, China, lzh_cathy@sjtu.edu.cn, Yongpei Guan, Guohua Wan A mixed formulation is presented for a real world planning and scheduling instance with multistage, multipurpose batch plants in rolling horizon. Besides, the input of storage between mixing and filling is discrete while the output is continuous. Sequence dependent setup and storage expiration are also taken into consideration. Lot sizing is achieved from an aggregation planning model for management’s sake, based on which a schedule model is formulated to get the optimal scheduling. Simulation is conducted to find the bottleneck of this procedure. 2 - Optimization of Automated Guided Vehicle Dispatching in Shop Floors Shiyang Huang, Iowa State University, 0076 Black Engineering, Ames, IA, 50011, United States, shuang@iastate.edu, Guiping Hu Automated guided vehicle (AGV) have been widely adopted for material handling on shop floors. We proposed a series of strategies to guide the dispatching of AGV based on heuristic rules and optimization models of material handling on shop floors. The optimization models were formulated and specially solved with designed algorithms that were proved to be more efficient than common solution techniques. The strategies were demonstrated in a case study based on a shop floor in literature, and the simulation results showed that the proposed methods outperformed classic strategy in optimizing different criteria on the shop floor. 3 - A New Hybrid Supply Chain Strategy with Expected Demand in Regional Facilities and a Common Service Plant Sanam Azadiamin, Ohio University, 35 N. Mckinley Ave. Apt 112, Athens, OH, 45701, United States, sa349416@ohio.edu, Gursel Suer A hybrid global supply chain strategy is proposed where regional plants are designed to meet expected demand whereas one service plant covers the remaining demand in all regions. This strategy requires transportation lead times and safety stocks to be considered to achieve a certain service level at a minimum cost. The relation between service levels and safety stocks is explored. 351A Production and Scheduling Contributed Session Chair: Feifan Wang, Arizona State University, Tempe, AZ, United States, feifan.wang@asu.edu 1 - Assembly Line Balancing: the Difficulty of Solving Certain Instances Jorge Pereira Gude, Universidad Adolfo Ibanez, 750 Avda Padre Hurtado, Office 216-C, Vina del Mar, 2530852, Chile, jorge.pereira@uai.cl The simple assembly line balancing problem is a classical optimization problem, which has been widely studied in the literature. While recent effort has been geared towards the identification of difficult instances, the question of what makes them difficult is still open. In this work, several new exact solutions WB29 WB31

procedures are put forward and their behavior under different instance characteristics is considered. An extensive computational experiment with these methods allows us to close several open instances and to give some insight into the difficulty of solving certain instances. 2 - Multi-job Production Systems: Theory and Application Pooya Alavian, PhD Candidate, University of michigan, Ann Arbor, MI, 48109, United States, alavian@umich.edu, Semyon M. Meerkov Multi-job production (MJP) is a class of flexible manufacturing systems intended to produce different job types based on a product-mix or build-schedule.In MJP, jobs are processed by the same sequence of operations, but with different processing times for each job type.We introduce the Work-based model, use it to characterize MJP systems, analyze the performance of MJP systems as functions of the product-mix, and report an application in automotive industry. 3 - Flexible Maintenance Schedules for Multistage Degradation Systems Yunyi Kang, Arizona State University, 1850 E Hayden Lane, Apt 207, Tempe, AZ, 85281, United States, ykang37@asu.edu, Feng Ju We present a flexible preventative maintenance problem on two-machine-and- one-buffer manufacturing systems, in which machines are subject to multi-stage degradation. Condition-based maintenance decisions are made on both machines, considering machine degradation stages and the buffer level. Moreover, flexible policies are included such that a machine can be recovered to any better operating state, comparing to which to only the best operating state is possible in many previous work. A Markov decision model is formulated to find the optimal maintenance policy in finite production runs and computational experiments are designed to show the improvements comparing to other policies. 4 - Hierarchy Machine Setup for Lot Scheduling at Assembly and Test Facilities Youngbum Hur, PhD Candidate, University of Texas at Austin, 11500 Jollyville Road, #1917, Austin, TX, 78759, United States, Youngbum.hur@utexas.edu, Jonathan F.Bard The importance of machine setups has been neglected in scheduling assembly and test (AT) operations in semiconductor back-end facilities. We examine the full dimensions of the problem in a multi-machine, multi-tooling environment with minimize the number of demand shortages and maximize weighted throughput. Our model determines machine setups, lot assignments and lot sequences using a greedy randomized adaptive search procedure (GRASP). In this paper, we consider a hierarchical approach to scheduling lots and to investigate the impact of different setup times and hierarchy rule. The performance of methodology is evaluated using data provided by a semiconductor company. 5 - An Integrated Approach to Production Planning and Scheduling for Automotive Wiring Harness Cut and Crimping Lines Jeongin Koo, Korea Institute of Industrial Technology, 89, Yangdaegiro-gil, Ipjang-myeon, Seobuk-gu, Researcher Bldg, This study considers a production scheduling problem of automotive wiring harness cut and crimping lines where all the harnesses should be processed manually or automatically by different types of machines. To produce wiring harnesses, every order of wiring harness production should be handled by the process allocation processes, the equipment allocation processes, and equipment scheduling in the cut and crimping lines. Given allocated processes for each wire to produce, we formulate an mixed integer linear programming model of the equipment allocation and scheduling to solve the case study of the wiring harness production line. 6 - Real Time Control of Geometric Serial Lines with Residence Time Constraints Feifan Wang, Arizona State University, 669 S.Mill Ave, Tempe, AZ, 85281, United States, feifan.wang@asu.edu, Feng Ju Scrap due to residence time constraints are commonly seen in manufacturing systems. In such systems, minimum and maximum waiting times for each part are enforced in the buffer. To optimize the trade-off between production rate and scrap rate, an on-off control policy for a two-machine geometric production system is derived using Markov decision processes with state approximation. An iteration method is further provided to improve the control policy. Cheonan, 31056, Korea, Republic of, jikoo@kitech.re.kr, Minjae Ko, Hwanju Ryu, Sangsik Ha, Hyunchul Tae

484