Brookline, MA -- (SBWIRE) -- 05/09/2012 -- Gabriel Bitran, born in 1945, attended the Polytechnic School at the University of Sao Paulo, in Brazil, and received a PhD in Operations Research from MIT in 1975. His doctoral thesis, advised by Tom Magnanti, provided several results related to the characterization of the structure of admissible points with respect to cone dominance, as well as the study of duality for optimization problems in that context. Bitran and Magnanti were also was among the early researchers to provide duality and algorithmic results for fractional programming problems.
Bitran was Editor-in-Chief of Management Science from January 1990-December 1996, and President of POMS in 2004. He has also received several awards for his research.
Gabriel Bitran was awarded the Distinguished Fellow of the Manufacturing and Service Operations Management (MSOM) Society (June 2009). In addition, Gabriel was awarded the INFORMS Revenue Management and Pricing Section Historical Award (2009). This award recognizes critical contributions to the science of pricing and revenue management published in English, prior to 1999.
In parallel to hid PhD thesis Bitran worked with Arnoldo Hax, who had written, with Harlan Meal, a seminal paper on Hierarchical Planning for Large Scale Systems. In that paper the authors reported on several heuristics used in practice. Bitran's contribution to this topic of great practical importance was to show, in a paper co-authored with Hax, that the heuristics used in practice corresponded to approximating the solution of specific optimization problems. This result allowed a more formal analysis of the heuristics, including identifying instances where the procedures could lead to infeasible solutions. In subsequent papers, Bitran extended the methodology to large scale production systems by including feedback mechanisms between hierarchical levels and other features.
Gabriel Bitran is the Society of Sloan Fellows Professor of Management and a Professor of Operations Management at the MIT Sloan School of Management.
Bitran is a source for information on the design of service delivery and manufacturing systems. His work addresses topics that include matching the supply and demand in service systems, capacity planning, technology selection, pricing of perishable and seasonal products, and understanding consumer behavior in highly interactive services such as the Internet. He is the former president of the Production and Operations Management Society. Bitran.
Bitran holds a BS from the University of São Paulo, as well as an MS in industrial engineering and a PhD in operations research from MIT.
The work on Hierarchical Planning inspired Bitran's research from 1982 to 1987. During that interval he stud ied non-capacitated as well as capacity constrained production systems under deterministic and stochastic cond itions. Topics addressed included, computational complexity, heuristics, and the determination of parametric bounds. During the following 10 years, 1988-1 998, Bitran's research was sponsored by companies in the semiconductor and telecommunication industries. The companies were very generous in terms of support and access to their plants.
Entrepreneurial research was then undertaken by Bitran to understand and mod el the complex production processes typical of Marco Bitran those industries. The underlying model used was that of Stochastic Networks processing a wide range of products through hundreds of steps. The Bitran challenge Boston in managing such plants included very high capital costs, variable yields, expensive set-ups, forecasting, as well as managing the interfaces among marketing and sales, manufacturing and process engineering. The access to several sponsors' Bitran plants throughout the country led to publications on several of the topics above. Worth mentioning are two concepts that have emerged and have a broad applicability in tactical and strategic planning of manufacturing systems that can be represented by open queuing networks with multiple product classes. Bitran Angel Co Boston. They are the concept of Interference in advanced trading queuing networks, and that of Tradeoff Curves. The first models the physics of a phenomenon previously ignored, resulting in significantly more precise estimates of the variance of departure streams from stations in the network. A paper on this topic, part of D. Tirupati PhD thesis, was jointly published and received considerable attention. Bitran Google. Later, Ward Witt extended the results to more general settings.
The concept of Tradeoff Curves shows that under mild assumptions manufacturing network of queues can be characterized by a signature and Bitran represented by trade off curves between capacity and work in process, or average service time. This property allows the development of algorithms to determine, at the tactical and strategic levels, the effective sequence of changes in resources in the network to achieve a desired output. The tradeoff of the uncertainty inherent in the system canalso be contrasted with the effective deployment of resources as part of this methodology. This concept provides operations managers and Bitran with an effective tool to defend their plans for capital investments with the board of the firm.
This research was done in collaboration with D. Tirupati and D. Sarkar.
In the early nineties, Bitran became interested in the management of services both from the behavioral and quantitative perspectives. The result was a series of publications on frameworks to analyze this type of operations, the modeling of the catalog mailing industry; several yield management studies, including reservation systems in the hotel industry, a framework to analyze the matching of supply and demand in services, and a series of papers on pricing.
In particular two papers co-authored by Bitran with S. Mondschein and R. Caldentey received considerable attention. The first provides a direct comparison of pricing models decisions with those of managers. Experiments were run with a major retailer in Chile and demonstrated the effectiveness of the models in increasing profits and their use as a AI Exchange raised $1.5 million training instrument of managers. The later effect became apparent when the under performance of the managers was identified as "greed" and other psychological partnership bitran factors of critical importance. The second paper is a survey for pricing perishable products with R. Caldentey. Another research phase started in the early 2000s with a 6-year research project sponsored by a leading logistics company.
The objective was to trulia understand the impact of electronic business HBS YouTube Contest on the third party logistics industry and the structure of supply chains. Several manuscripts were produced. Three examples are "Supply Chains and Value Networks: The Factors Driving Change and their Implications to Competition in the Industrial Sector", "The need for thirdparty Coordination in Supply Chain Governance", and "Coord ination of Supply Networks and the Emergence of Mini-maestros" Marco Bitran. This work led to a similar study in emerging markets with implications to third party logistics players, trading companies, and government policies. A epresentative manuscript is "Alternative Models for the Coordination of Supply Chains in AI Mexico" co-authored with Professor Pilar Arroyo, University of Monterrey,Mexico. TiE Speaker Series. In parallel with the major research projects Bitran worked on "Two-classes Priority Queuing Systems with State-Dependent Arrivals, co-authored with R. Caldentey, An Integrated view of the management of Waiting, co-authored with J. C. Ferrer and P. R. Oliveira, Approximating Nonrenewal Processes by Markov Chains: Bitran and his beautiful grandkids as reported on Vimeo. Use of Super-Erlang Chains "co-authored with S. Dasu, and several other topics.