Wagner Prize Presentations

 History and Purpose of the Wagner Prize and Competition

Presented by Allen Butler, President, Daniel H. Wagner Associates, Inc., 2 Eaton Street, Suite 500, Hampton, VA, 23669, United States of America, allen.butler@va.wagner.com.

• View Presentation

 Approximate Dynamic Programming Captures Fleet Operations for Schneider National

Presenting Author: Hugo Simao, Research Staff, Princeton University - CASTLE Lab, Sherrerd Hall 112, Princeton NJ 08544, United States of America, hpsimao@princeton.edu

Co-Authors: Jeff Day, Schneider National; Abraham George, Princeton University, Dept of Operations Research and Fin Engg, Princeton, United States of America, abegeorge@gmail.com; Ted Gifford, Schneider National, GiffordT@schneider.com; John Nienow, Schneider National, nienowj@schneider.com; Warren Powell, Professor, Princeton University, Sherrerd Hall 230, Princeton NJ 08544, United States of America, powell@princeton.edu

Abstract: Schneider National needed a model that would replicate the behavior of their team of dispatchers. We used the modeling and algorithmic framework of approximate dynamic programming to optimize the movements of 6,000 drivers, each described by a 15 dimensional attribute vector, over a month. The model closely replicates historical performance, and also produces accurate estimates of the marginal value of each of 500 types of drivers. Numerous projects have produced millions in savings.

• View Presentation

 A Queuing Model-Based System for Semiconductor Production Planning at IBM

Presenting Author: Horst Zisgen, IBM Deutschland Research and Development GmbH, Hechtsheimer Strasse 2, Mainz 55131, Germany, horst_zisgen@de.ibm.com

Co-Authors: Steven M. Brown, IBM Systems and Technology Group, Hopewell Junction NY 12533-6683; Thomas Hanschke, Department of Mathematics, Clausthal University of Techology, Erzstrasse 1, Clausthal-Zellerfeld 38678, Germany; Ingo Meents, IBM Deutschland Research and Development GmbH, Hechtsheimer Strasse 2, Mainz 55131, Germany; Benjamin R. Wheeler, LGO Fellow at MIT Sloan School of Management, 1 Amherst Street Building E40-315, Cambridge MA 02142

Abstract: This paper presents IBM's Enterprise Production planning Optimization System (EPOS). EPOS uses a unique queuing network model combined with linear programming to meet the requirements of capacity and lead time planning for semiconductor manufacturing. EPOS is in continuous use at IBM's 300mm factory where it has become an integral tool for predicting bottlenecks, prioritizing continuous improvement efforts, planning capital investment, and managing factory lead times. EPOS has guided efforts to improve factory performance and reduce more than 30 million dollars of expense.

• View Presentation

 Hub Group Implements a Suite of OR Tools to Improve its Operations

Presenting Author: Michael Gorman, Associate Professor, University of Dayton, 300 College Park, Dayton OH 45469-2130, United States of America, michael.gorman@udayton.edu

Abstract: Hub Group developed a production decision support system which integrated forecasting, error distribution analysis, expected value based heuristics and optimization. The system was structured to fit within Hub's existing organizational structure, business processes and information technology to reduce project risk, cost and operational disruption. Hub improved its performance by $11 million in cost savings in the system's first year, improved revenue per load and increased container velocity.

• View Presentation

 Extending Bass for Improved New Product Forecasting at Intel

Presenting Author: Karl Kempf, Intel Corporation, 5000 W. Chandler Blvd., MS-CH3-10, Chandler AZ 85226, United States of America, karl.g.kempf@intel.com

Co-Authors: Mehmet O. Atan, Lehigh University; Berrin Aytac, Lehigh University, 200 W. Packer Ave., Bethlehem PA 18015, United States of America, bea4@lehigh.edu; Asima Mishra, Research Fellow, Intel Corporation; Shamin A. Shirodkar, Intel Corporation; S. David Wu, Lehigh University

Abstract: Forecasting demand for new products is increasingly difficult as the technology treadmill drives product lifecycles shorter and shorter. We present a model that perpetually reduces forecast variance as new market information is acquired over time. Our model extends Bass's idea of product diffusion to a more comprehensive theoretical setting using the notion of demand-leading indicators in a Bayesian framework. Successful implementation at Intel demonstrates not only improvement in time/efforts but also reduction in forecast errors that leads to significant cost savings.

• View Presentation

 Optimizing Helicopter Transport of Oil Rig Crews at Petrobras

Presenting Author: Hernan Abeledo, Associate Professor, The George Washington University, 1776 G St NW, Washington DC 20052, United States of America, abeledo@gwu.edu

Co-Author: Marcus Poggi de Aragao, Departamento de Informatica, Pontificia Universidade Catolica Rua Marquez de Sao Vicente 225, Rio de Janeiro RJ 22451-9, Brazil, poggi@inf.puc-rio.br; Fernanda Menezes, Gapso Tecnologia da Decisao Rua Lauro Muller, 116 / 3402, Rio de Janeiro RJ 22290-, Brazil, fmenezes@gapso.com.br; Lorenza Moreno, Departamento de Informatica, Pontificia Universidade Catolica Rua Marquez de Sao Vicente 225, Rio de Janeiro RJ 22451-9, Brazil, lorenza@inf.puc-rio.br; Nelci Nascimento,Exploracao e Producao - Servicos / US-TA Av. Rui Barbosa, 1940, Macae, RJ 27.915-, Brazil, nelcar.classic@petrobras.com.br; Oscar Porto, Gapso Tecnologia da Decisao Rua Lauro Muller, 116 / 3402, Rio de Janeiro RJ 22290-1, Brazil, oscar@gapso.com.br; Marcelo Reis, Gapso Tecnologia da Decisao Rua Lauro Muller, 116 / 3402, Rio de Janeiro RJ 22290-1, Brazil, marcelo@gapso.com.br; Eduardo Uchoa, Universidade Federal Fluminense, R. Passo da Patria 156, Engenharia de Produasaeo, Niteroi RJ 24210, Brazil, uchoa@producao.uff.br

Abstract: Petrobras carries over 1900 workers daily by helicopter to about 80 offshore oil platforms. This operation is among the largest worldwide for civilian helicopters. We present a flight scheduling system designed to improve service and minimize cost. Achievements comprise strict enforcement of safety guidelines, reductions of 18% in number of offshore landings, 8% in total flight time, and 14% in flight costs. The optimization model, involving routing and packing, is solved by column generation.

• View Presentation