2016 INFORMS Annual Meeting Program

MC34

INFORMS Nashville – 2016

3 - A Multistart Algorithm For The Parallel Machine Scheduling Problem With Dependent Setup Times And Preventive Maintenance Oliver Avalos-Rosales, Profesor Investigador de Tiempo Completo, Universidad Autónoma de Coahuila, orion 338, Satelite Norte, Saltillo, 25115, Mexico, aoliver84@gmail.com, Ada M. Álvarez, Francisco Angel-Bello We address an unrelated parallel machine scheduling problem minimizing the makespan. We consider dependent setup times and periodic preventive maintenance. These aspects have not been jointly studied in parallel machine environment. The problem is NP-hard. We consider the structure of feasible solutions and the structure of the objective function to design each component of the multistart proposed algorithm. We present computational experiments to compare the algorithm with exact solutions in small an medium size instances, and validate the contribution of each part of the algorithm in large instances. hamed.yarmand@umb.edu, Amirreza Shojaeifard, Babak Rezaee We present a novel model for the elective surgery scheduling problem for multiple operating rooms to improve the efficiency of ORs with the intent of maximizing the profit (considering revenue of surgeries, fixed cost, and overtime cost). We develop an integer linear programming model for this scheduling problem. The developed model is a four dimensional assignment problem that determines the weekly schedule (day, surgeon, operating room, and type of surgery) considering three types of surgery demands simultaneously (pre- scheduled, pre-assigned, and other). It also considers surgeons’ availabilities for performing surgeries. Two heuristic algorithms are proposed and investigated. 5 - Project Planning And Scheduling To Maximize Expected Quality In The Presence Of Stochastic Time Delays Matthew J Liberatore, Villanova University, 800 Lancaster Avenue, Villanova, PA, 19085, United States, matthew.liberatore@villanova.edu, Bruce Pollack-Johnson We present research designed to help deal with probabilistic time delays and cost overruns which endanger project quality. We present strategies to find the planned durations for tasks and the scheduling protocol that maximize the expected overall project quality by applying our previously developed notion of a continuous quality function for a task (and the project overall) in terms of the time and investment put into it. 4 - Operating Room Scheduling Under Hybrid Demand Hamed Yarmand, University of Massachusetts Boston, 2 Fatima Rd, Stoneham, MA, 02180, United States, Chair: Barret Pengyuan Shao, Crabel Capital Management, 414 East Market Street Second Floor, Charlottesville, VA, 22903, United States, barretshao@gmail.com 1 - Simulation-based Optimization For Layout-based Grafting Resource Allocation Sara Masoud, University of Arizona, 1300 E Fort Lowell Road, # A214, Tucson, AZ, 85719, United States, saramasoud@email.arizona.edu, Young-Jun Son, Chieri Kubota, Russell Tronstad Optimal resource planning in vegetable seedling propagation facilities is complicated due to the dynamicity of workers’ performance. In addition, the negative impact of an inefficient layout on workers’ performance reduces the productivity in grafting facilities substantially. In this work, a simulation-based optimization model is devised to achieve the optimal layout-based resource allocation. The proposed model is customized based on workers’ individual performance and managerial design preferences. The optimal solution is expected to reduce the production cost of grafting systems. systems. 2 - Interfirm Imitation Under Relational And Institutional Influences Kyun Kim, Doctoral Student, University of Texas at Dallas, 800 West Campbell Road, SM43, Richardson, TX, 75080, United States, kyun.kim@utdallas.edu, Zhiang (John) Lin In the interfirm imitation research, the role of imitation targets is often underexplored since imitators have received the most attention. Also, the connection between macro and micro constructs related to imitation has not been clearly discussed. We endeavor to shed light on imitation targets and to connect macro and micro factors. We develop a status based approach and introduce how status (macro and micro) of a firm lets imitators reduce uncertainty and gain legitimacy. We also examine performance (micro) and institutional environment (macro) of an imitation target. We conduct empirical tests and computational analyses regarding firms’ resource acquisition activities: M&As and Alliances. MC33 203B-MCC Simulation I Contributed Session

3 - Climate Prediction Markets And Investor Beliefs: An Agent-based Simulation

Jonathan Gilligan, Associate Professor, Vanderbilt University, 2301 Vanderbilt Place, PMB 351805, Nashville, TN, 37235-1805, United States, jonathan.gilligan@vanderbilt.edu, John J. Nay, Martin Van der Linden A large fraction of the American public doubts the scientific consensus that human activity is causing global warming. Climate prediction markets might influence beliefs in people who distrust scientists but trust free markets. We present an agent-based computational test-bed to examine prediction market dynamics. Traders with different beliefs about climate bet on future temperatures and adapt their beliefs based on the profits of other traders. Traders believe that global climate is primarily controlled by carbon dioxide or by solar irradiance. Market participation causes traders’ beliefs to converge rapidly, suggesting that a climate market could be useful for building public consensus. 4 - Agent-Based Simulation Of Production And Seeding Strategies For Innovations We develop an agent-based simulation model of new technology diffusion to evaluate different viral marketing and inventory build-up policies under various social network structures. We show that determining seeding and build-up policies sequentially may lead to suboptimal decisions. We show how the optimal joint policy varies for different product categories and that the seeding strategy that maximizes demand rate is not necessarily optimal under supply constraints. We also investigate the role of high-degree nodes and long-range connections in scale-free and small-world networks. 5 - Approximation Of Long Memory Process With Short Memory Process And Some Numerical Experiments Barret Pengyuan Shao, Crabel Capital Management, 414 East Market Street Second Floor, Charlottesville, VA, 22903, United States, barretshao@gmail.com Taking FARIMA(p,d,q) process with d > 0 as an example for long memory process, we use information distance to prove that stationary ARMA processes are dense in all FARIMA processes in the total variation distance. As a consequence, statistical tests with finite sample size fail to distinguish a FARIMA process from ARMA processes. We provide Monte Carlo experiments that confirm that long memory processes are not easily distinguished from our approximate ARMA processes with finite sample sizes using a variety of well known statistical tests. 204-MCC Empirical Studies in Healthcare OM Sponsored: Manufacturing & Service Oper Mgmt, Healthcare Operations Sponsored Session Chair: Song-Hee Kim, USC Marshall School of Business, 3670 Trousdale Parkway, Los Angeles, CA, 90089, United States, songheek@marshall.usc.edu 1 - Nursing Shift Assignment And Its Influence On Medical Outcomes: First Insights Of A Multicenter Study Ashkan Negahban, Assistant Professor, Pennsylvania State University, 30 E Swedesford Rd, Malvern, PA, 19355, United States, anegahban@psu.edu, Jeffrey Smith MC34 Low staffing levels are known to be a risk factor for medical outcomes. However, it is important not only to consider the right staffing levels but also to assign the available staff in the most sensible manner. Based on data from a multicenter study of 66 neonatal intensive care units, we analyze variation in staffing allocation decisions and present the first insights into their association with outcomes. 2 - The Impact Of Internal Service Quality On Nurse Inefficiency And Medical Errors Xin(Sarah) Zheng, Boston University, Boston, MA, United States, xinzheng@bu.edu, Anita L Tucker, Z Justin Ren, Janelle Heineke, Amy McLaughlin, Aubrey Podell Drawing on the theories of swift, even flow and conservation of resources, we propose a new avenue for addressing medical errors—improving internal service quality (ISQ), which is the quality of service provided by support departments such as housekeeping, and materials management. Using 13 months of panel data from five nursing units that gather weekly data on ISQ delivered by 11 support departments, we find that a one standard deviation increase in ISQ is associated with near elimination of hospital-acquired pressure ulcer and patient fall. The one standard deviation increase in ISQ is further associated with a 5% reduction in nurse inefficiency and a financial benefit as high as $7 million. Ludwig Kuntz, University of Cologne, Cologne, Germany, kuntz@wiso.uni-koeln.de, Felix Miedaner, Stefan Scholtes

193