Informs Annual Meeting Phoenix 2018

INFORMS Phoenix – 2018

POSTER SESSION

12 - Sequencing and Line Balance Interaction in Multi-worker Mixed- model Assembly Lines Mary Beth Kurz, Clemson University, 110 Freeman Hall, Clemson, SC, 29634-0920, United States Mixed-model assembly requires line balancing and sequencing. The line balance and sequence may induce work overload. Given a fixed line balance, a sequence can be adjusted to minimize work overload: a skip policy allows a utility worker to take over a workpiece that will cause overload while the normal worker works on the next workpiece and swimming allows a worker to move into adjacent stations. Large products are assembled at stations with one or more workers per station. We address sequencing with skips and swims in a multi-worker mixed- model assembly line. 13 - A Cellular Manufacturing Implementation in Microwave Manufacturing Okan Karatay, Production Manager, Aselsan, Konya Yolu 8. Km, Ogulbey Mah. 3051. Sok. No:3, Ankara, 06830, Turkey, Yuce Cinar Baysal Nowadays, one of the effective ways of dealing with manufacturing drawbacks in a high mix low volume environment is the CM. In this study, we first propose a manufacturing cell that is designed to replace the traditional job shop layout of the Microwave manufacturing company. Secondly, we focus on the scheduling that creates the continuous flow and provides the line balance. Moreover, we mention the human aspect of this implementation which is as critical as technical ones. Thirdly, we evaluate the performance of the CM with the conventional production layout by using the companys KPI. After analyzing the outcomes of the implementation, the benefits associated with cellular manufacturing are listed. 14 - A Bayesian Stochastic Frontier Model for Productivity Index Jessie Nouri, UWM, Milwaukee, WI, United States Stochastic Frontier Model (SFM) is a linear regression with covariates and a random effect term which captures variation of production across various firms. This research develops a semi-parametric Bayesian SFM for predicting the productivity index across production lines of a major firm. The distribution of efficiencies is modelled through a Dirichlet process prior. 15 - The Relation of Product Variety and Structure to Subassembly Inventory Jeonghan Ko, Ajou University; University of Michigan, Ann Arbor, MI, 48109, United States, Kunju Kim This research elaborates the relationship between product variety and the safety inventory of subassemblies thorough multi-stage assembly. We identifies the optimal locations to hold subassemblies along the assembly steps to minimize the total subassembly inventory levels on the condition of little component commonality. In particular, we show that different product structure will lead to different levels and optimal locations for subassembly inventory. 15B - Real-time Health Monitoring of Wiring Systems for Smart Manufacturing Daeil Kwon, Professor, UNIST, Ulsan, 44919, Korea, Republic of, Jinwoo Lee, Namhun Kim In manufacturing facilities, wiring systems transmit and receive signals to communicate between modules and systems. Stress conditions may degrade the performance of wiring systems, causing intermittent faults, interlocking, and eventually facility downtime. Timely maintenance of wiring systems is required to ensure the reliability and availability of manufacturing facilities.This presentation discusses a non-destructive method to allow for condition-based maintenance of wiring systems. Experimental results demonstrate the proposed method can serve as a real-time health indicator of wiring systems, and can be used for proactive maintenance planning toward near zero downtime. 16 - What Effects do Owner-operator Incur as a Result of Negotiating Behavior for Freight Transportation Shervin Espahbod, Wilfrid Laurier University, Waterloo, ON, Canada The main approach to be a self-employed in trucking industry is to purchase or lease a truck as an independent trucker. The self-employed driver can receive direct orders or subcontract for a transportation firm. Although there is a substantial potential profit for self-employed carriers, the trucking industry is an extremely competitive and uncertain market. The person with the lowest amount of bid, is the winner of the spot market auction. By cutting their profits and putting lowest bids, self-employed drivers win bids from far distance, while closer corporate truckers are unable to win the bid. This study will examine the results of the inefficiency from this behavior.

17 - Do Attractive People Make a Better Deal Investigation of Social Factors in Supply Chain Management Lyudmyla Starostyuk, PhD Candidate, The University of Texas at Arlington, 701 S. West Street, Arlington, TX, 76019, United States, Chen Kay-Yut, Prater Edmund L Although it is well known that people form impressions of others based on their appearance, little is known about the role of facial character judgments in contracting. We study the role of social factors such as trustworthiness, attractiveness, and dominance in better pricing in a supply chain contract game. The experimental study shows that the female supplier who is confident in her attractiveness sells products to the male buyer for the higher average price while the male supplier does the opposite. The trustworthy looking female receives more expensive offers from the male supplier but sells products cheaper to the female buyer as well as the dominant female does. 18 - The Importance of Referent Power in Team Mental Models Shihao Ge, Research Assistant, University of Washington Bothell, 18115 Campus Way NE, Bothell, WA, 98011, United States, Deanna Kennedy Being on the same page by forming team mental models is important for project success. This takes communication, however, the communicator may help or hinder the development of team mental models. Herein, we look at how relationships and behaviors of the communicator on other team members and how that may impact mental model formation. 19 - Predicting Hurricane Landfall from the Atlantic on United States Territories Using Historical Advisory Data Obe J. Joseph, SUNY University at Buffalo, Buffalo, NY, United States The objective of this study is to create a landing deviation model. The study involves collecting and tabulating advisory text data from 1998 to 2017, from the National Hurricane Center. It analyzes the distribution of the distance deviation between the forecasted landfall and the true landing, given the number of hours before landfall. 20 - Natural Disaster Damage Loss Prediction Using Social Media Data: The Study Case of Hurricane Irma Ling Lin, University of Buffalo, 319 Kaymar Dr, Buffalo, NY, 14228, United States, Qing He, Rajan Batta Our research presents an aggregated disaster loss prediction model by examining Hurricane Irma’s Tweet data and users’ demographic information through implementing sentiment and linear regression analyses. Sentiment scores were utilized as new input variables along with other factors in regression model to predict various types of disaster damage losses. The results show the potential power of social media data in predicting both the number of damaged units and the damage loss in dollars. 21 - Optimization of Information Coverage Using Social Media During Natural Disasters Bairong Wang, PhD, University at Buffalo (SUNY),Buffalo, NY, 14226, United States, Rajan Batta One of the challenging problems for social media crisis sensing is collecting enough useful data from social media. This study aims at designing and optimizing an information coverage strategy within local areas when a disaster event occurs. A bi-level information coverage model is designed to model the situations where information is collected from different local areas for crisis sensing. This model includes an upper-level model that optimizes the resources allocating among different local areas and a correspondingly lower-level model that optimizes the positioning of computing resources to maximize the weighted coverage. Test results of the model prove a sound performance of the model. 22 - Using Real-time Operational Data to Increase Labor Productivity in Retail This research is focused on developing models to support dynamic systems that can reassign labor to different activities. This work utilizes data from a pilot study in a store of a home improvement retail chain in the U.S. Security cameras were used to track the number of customers and employees on different sections. We use these operational data to measure - in real time - the demand for services in a specific store department, and the service capacity. This information can be linked to point of sales data, to measure the effect of the number of employees on the sales of specific departments. All this provide information for a dynamic labor allocation system in order to prioritize where to allocate employees. 23 - Strategies with Incomplete Information for Interactions Between Pablo Jofr , Universidad de Chile, Beaucheff 851, Santiago, 8370456, Chile, Marcelo Olivares, Andres I. Musalem This work studies the intention of private sector investment under tenuous rules of law (Incomplete Information) and associated with concession period length. First, a static game model was created to analyze the tactical concession length, indicating 60% of the legal maximum length. Then, to analyze private investment promotion, another model was created to investigate the influences of the investment ratio. Results from the models show: 1) “zero concession is negated; 2) only under extreme risk the private sector’s motivation will be influenced, but with limitation. Sectors in Concession Period Length of a PPP Project Zhouyang Lyu, Hohai University, Nanjing, 211100, China

178