2016 INFORMS Annual Meeting Program

MC52

INFORMS Nashville – 2016

MC52

MC54 Music Row 2- Omni Service Innovation in the Cognitive Era Invited: Service Science Invited Session Chair: Changrui Ren, IBM Research - China, IBM Research - China, Beijing, 100193, China, rencr@cn.ibm.com 1 - Enterprise Cloud Garbage Collector Sai Zeng, IBM T.J. Watson Research Center, Yorktown Heights, NY, United States, saizeng@us.ibm.com, Christopher Young, Karin Murthy Infrastructure as a Service (IaaS) clouds empowers the agility to provision servers. Recent findings indicate that this agility led to a situation where 1 in 3 data center servers is a zombie server, aka server is running but does not do any useful work. In this paper, we present Enterprise Cloud Garbage Collector, a tool that detects zombie severs. It establishes dependency between users/clients and servers by constructing a weighted reference model based on application knowledge. In the situation of insufficient application knowledge, it supplements its dependency results with a machine learning model trained on resource utilization data. 2 - Big Data Fueled Supply Risk Management: Sensing, Prediction, Evaluation And Mitigation Changrui Ren, IBM Research - China, Beijing, China, rencr@cn.ibm.com, Miao He, Qinhua Wang Supplier risks jeopardize on-time or complete delivery of supply in a supply chain. This talk will introduce a big data fueled approach to monitor and manage supply risks, which includes a big data analytics component, a simulation component and an optimization component. The big data analytics component senses and predicts supply disruptions with internally (operational) and external (environmental) data. The simulation component supports risk evaluation to convert predicted risk severity to key performance indices (KPIs) such as cost and stockout percentage. The optimization component assists the risk-hedging decision-making. Chair: Heinrich Kuhn, Catholic University of Eichstaett-Ingolstadt, Auf der Schanz 49, Ingolstadt, 85049, Germany, heinrich.kuhn@ku.de 1 - Base-stock Models For Lost Sales - A Markovian Approach Sang-Phil Kim, Assistant Professor, Winona State University, 175 W. Mark st., Somsen 406, Winona, MN, 55987, United States, Ksphil@me.com, Yanyi Xu, Maqbool Dada, Arnab Bisi, Suresh Chand We consider the lost sales model with discrete demand. The inventory is reviewed every T periods and an order is placed to bring the inventory position back to the target base-stock level R, and is received after a lead time of L periods. Based on the outstanding orders in the pipeline, we represent the state of the system as a Markov chain. We show that the structure of the transition probability matrix is recursive in R and L. This special structure is used to facilitate computation of the stationary distribution. Analytical results complemented by numerical examples reveal that neither the optimal base-stock nor the expected cost is monotone in L for a given T. 2 - Capacity Usage Estimation Methodology For Inventory Management Ahmet Nuroglu, Yildiz Technical University, Barbaros Bulvari, Yildiz-Istanbul, 34349, Turkey, envernuroglu@gmail.com, Fahrettin Eldemir New analytical capacity usage estimation methodology for economic order quantity (EOQ) model is proposed. In multiple item warehouse-space capacity constrained EOQ model, by applying the randomized storage concept, capacity usage is estimated from expected inventory occurrences instead of order quantities. In joint replenishment problem under power of two (PoT) policy, the capacity usage is estimated from average inventory occurrences which are the function of PoT parameter of each item. The feasible optimal solutions are simulated and validated. MC55 Music Row 3- Omni Inventory Management II Contributed Session

214-MCC Panel: Pro Bono Analytics Sponsored: Public Sector OR Sponsored Session

Moderator: David T. Hunt, Oliver Wyman, One University Square, Princeton, NJ, 08540, United States, david.hunt@oliverwyman.com 1 - Pro Bono Analytics David T. Hunt, Oliver Wyman, One University Square, Princeton, NJ, 08540, United States, david.hunt@oliverwyman.com Pro Bono Analytics is an initiative within INFORMS to match members willing to volunteer their OR and analytical skills with non-profit organizations working in underserved and developing communities. Panelists include Nashville area non- profit organizations and Pro Bono Analytics volunteers discussing how analytics can provide positive impacts for topics ranging from improving operations at a homeless shelter to understanding the inventory needs for supplies in a low- income school district. 2 - Panelist: Matthew Brondum, US Army Corps of Engineers, Vicksburg, MS, United States, mcb345@cornell.edu 3 - Panelist: Joel Wright, PENCIL Foundation, Nashville, TN, United States, jwright@pencilfd.org 4 - Panelist: Cindy Corona Rivera, Hands On Nashville, Nashville, TN, Cindy@hon.org 5 – Panelist: Anna Danandeh, Verizon, Waltham, MA, annadanandeh@mail.usf.edu Moderator: Nitin Joglekar, Boston University Questrom School of Business, 595 Commonwealth Avenue, Boston, MA, 02215, United States, joglekar@bu.edu 1 - Panelist: Nitin Joglekar, Boston University Questrom School of Business, joglekar@bu.edu This panel showcases alternative themes and research approaches being pursued by a select set of emerging scholars in the startup product, supply chain & technology management research domain. 2 - Panelist: Jennifer Bailey, Babson College, jbailey@babson.edu 3 - Panelist: Jianxi Luo, Singapore University of Technology & Design, luo@sutd.edu.sg 4 - Panelist: Joel Wooten, University of South Carolina, joel.wooten@moore.sc.edu 5 - Panelist: Onesun Steve Yoo, University College London, onesun.yoo@ucl.ac.uk 6 - Panelist: Meyyappan Narayanan, Lakehead University, Thunder Bay, ON, Canada, meyyappan.narayanan@lakeheadu.ca MC53 Music Row 1- Omni Panel: Emerging Themes in Startup Product, Supply Chain & Technology Management Sponsored: Technology, Innovation Management & Entrepreneurship Sponsored Session

200