26.11.2012 Views

Technical Sessions – Monday July 11

Technical Sessions – Monday July 11

Technical Sessions – Monday July 11

SHOW MORE
SHOW LESS

You also want an ePaper? Increase the reach of your titles

YUMPU automatically turns print PDFs into web optimized ePapers that Google loves.

Mathematical models are used to explain why manufacturing systems evolved<br />

as they did over the last 300 years. I begin with the motives for the factory<br />

system, then move to the introduction of flow lines, the survival of batch production<br />

systems, the advent of moving belt assembly lines, transfer lines, group<br />

technology, flexible manufacturing systems and the Toyota sewing system. The<br />

talk will show how system improvement arose from an understanding of both<br />

variability and its sources and the characteristics of workers and machines.<br />

Some current challenges will be discussed.<br />

� TB-05<br />

Tuesday, <strong>11</strong>:00-12:30<br />

Meeting Room 104<br />

Marketing/Operations II<br />

Stream: OR and Marketing<br />

Invited session<br />

Chair: Rick So, UC Irvine, 92697, Irvine, CA, United States,<br />

rso@uci.edu<br />

Chair: Kumar Rajaram, UCLA Anderson School, 100 Westwood<br />

Plaza, 90095, Los Angeles, CA, United States,<br />

krajaram@anderson.ucla.edu<br />

1 - Distribution Planning to Optimize Profits in the Motion<br />

Picture Industry<br />

Kumar Rajaram, UCLA Anderson School, 100 Westwood Plaza,<br />

90095, Los Angeles, CA, United States,<br />

krajaram@anderson.ucla.edu<br />

We consider the distribution planning problem in the motion picture industry.<br />

This problem involves forecasting theater-level box office revenues for a given<br />

movie and using these forecasts to choose the best locations to screen a movie.<br />

We develop and test our methods on realistic box office data and show that it<br />

has the potential to significantly improve the distributor’s profits. We also develop<br />

some insights into why our methods outperform existing practice, which<br />

are crucial to their successful practical implementation.<br />

2 - Supply Chain Models with Preferred Retailer Privy to<br />

Supplier’s Inventory Information<br />

Hamed Mamani, Dept of Information Systems and Operations<br />

Management, University of Washington, Foster School of<br />

Business, 98195, Seattle, WA, United States,<br />

hmamani@u.washington.edu, Kamran Moinzadeh, Apurva Jain<br />

In a supply chain consisting of a supplier and several retailers, some of the<br />

retailers have preferred status, providing them with information about the supplier’s<br />

inventory level. Such status can be due to strategic behavior of retailers<br />

and learning the supplier’s replenishment policy. Preferred retailers, thus, can<br />

be proactive and inflate their orders when supply gets short. We study the dynamics<br />

of such supply chains as a Stackelberg game where retailers react after<br />

the supplier has fixed his strategy. We evaluate outcome of the resulting game<br />

with solution of the centralized problem.<br />

3 - The Planning of Guaranteed Targeted Display Advertising<br />

John Turner, Paul Merage School of Business, UC-Irvine, 92697,<br />

Irvine, CA, United States, john.turner@uci.edu<br />

Ad networks increasingly strive to apply planning models to broad classes of<br />

targeted display advertising, including banner ads, in-game advertising, and<br />

digital TV ads. We formulate a broadly-applicable ad planning problem as a<br />

quadratic transportation problem, and show that the quadratic objective is a<br />

good surrogate for several important performance metrics. Moreover, we exploit<br />

the quadratic objective and develop two complimentary algorithms which<br />

intelligently aggregate the audience space in a manner that dramatically reduces<br />

problem size while obtaining near-optimal solutions.<br />

4 - Diffusions of Mobile Cellular Phones in Sub Sahara<br />

Africa<br />

Chaiho Kim, OMIS Department, Santa Clara University, 500 El<br />

Camino Read, 95070, Santa Clara, CA, United States,<br />

ckim@scu.edu<br />

It is widely believed that mobile phones will play critically important roles for<br />

economic developments of countries in Sub Sahara Africa. Using statistical<br />

models, this paper will examine how the diffusions of mobile phones in Sub<br />

Sahara Africa are related to economic, cultural, and other variables such as different<br />

geographical regions. The paper will also examine whether the relations<br />

found for the mobile phones are significant different from those found for the<br />

main fixed phone lines.<br />

IFORS 20<strong>11</strong> - Melbourne TB-06<br />

� TB-06<br />

Tuesday, <strong>11</strong>:00-12:30<br />

Meeting Room 105<br />

Routing Problems :Innovative Applications<br />

and Solution Techniques<br />

Stream: Transportation<br />

Invited session<br />

Chair: Pierre Dejax, IRCCyN, Ecole des Mines de Nantes, La<br />

Chantrerie, 4, rue Alfred Kastler, BP 20722, 44307, Nantes cedex 3,<br />

France, Pierre.Dejax@emn.fr<br />

1 - The Split Delivery Capacitated Team Orienteering Problem<br />

M. Grazia Speranza, Dept. of Quantitative Methods, University<br />

of Brescia, C.da Santa Chiara, 50, 25122, Brescia, Italy,<br />

speranza@eco.unibs.it, Claudia Archetti, Nicola Bianchessi,<br />

Alain Hertz<br />

We study the capacitated team orienteering problem with split deliveries. Potential<br />

customers are associated with a demand and a profit. The customers to<br />

be served by a fleet of capacitated vehicles have to be identified to maximize<br />

the collected profit, with constraints on the maximum time duration of each<br />

route and the vehicle capacity constraints. We show that the profit collected by<br />

allowing split deliveries may be as large as twice the profit collected without<br />

split deliveries. We present exact and heuristic algorithms and test the proposed<br />

approaches on several sets of instances.<br />

2 - Flexible Buses for Manhattan<br />

Felisa Vazquez-Abad, Computer Science Dept., City University<br />

New York, 695 Park Ave, Room HN1000E, 10065, New York,<br />

United States, felisav@hunter.cuny.edu, Jennie Lioris, Guy<br />

Cohen<br />

Few privately owned cars circulate in Manhattan, yet the traffic congestion has<br />

become a serious problem. People with limited mobility and senior citizens<br />

avoid subways, cannot afford taxis and face excessive delays travelling by bus.<br />

We explore an alternative formulation where cars adjust their routes on demand,<br />

but may take several passengers. Pilot studies in Paris show that this system can<br />

achieve 80% occupation rate, instead of the 60% idle taxis roaming the streets<br />

today. We will present online routing allocation algorithms that use threshold<br />

optimization under different pricing models.<br />

3 - Enhanced Variable Neighborhood Search for Multi- Period<br />

Inventory Routing Problem Model with Time Varying<br />

Demand<br />

Noor Hasnah Moin, Institute of Mathematical Sciences,<br />

University of Malaya, 50603, Kuala Lumpur, Malaysia,<br />

noor_hasnah@um.edu.my<br />

We consider a multi-period Inventory Routing Problem (IRP) that faces time<br />

varying demand of multi-product from the assembly plant, in a many-to-one<br />

distribution network consisting of an assembly plant and many geographically<br />

dispersed suppliers each supplying a distinct product to the plant. The product<br />

is ready for collection when the vehicle arrives and all demand must be met<br />

without backlogging. The inventory holding cost is product specific and a fleet<br />

of capacitated homogeneous vehicles transport products from the suppliers to<br />

meet the demand of the plant in each period. The problem is formulated as a<br />

mixed integer programming problem. We propose a solution method based on<br />

the Variable Neighborhood Search where several heuristics are incorporated at<br />

various stages of the algorithm. The algorithms are run on several problems<br />

from the literature and the results are compared with the Genetic Algorithms.<br />

VNS performs better on larger problems.<br />

4 - Robust Scenario Generation for a Real Large-Scale Inventory<br />

Routing Problem<br />

Pierre Dejax, IRCCyN, Ecole des Mines de Nantes, La<br />

Chantrerie, 4, rue Alfred Kastler, BP 20722, 44307, Nantes<br />

41

Hooray! Your file is uploaded and ready to be published.

Saved successfully!

Ooh no, something went wrong!