Arthur F. Veinott Professor Management Science and Engineering Office: Terman 440 | Phone: 650-725-0548 | Fax: 650-723-1614 Email: veinott @ stanford.edu

Research Interests

Operations Research and Management Science

Dynamic Programming

Lattice Programming

Supply-Chain Optimization

Network Flows

Optimization

---

Dynamic Programming

This area currently focuses on the optimal management of populations over time, e.g., systems, firms, inventories, jobs, customers, labor force, vehicles, books, passengers, securities, fish, trees, patients. Interest centers on the effective and efficient choice of decisions in the presence of risk, sequential acquisition of information and possibly growth or decline of rewards and/or the population over time. The emphasis is on the development of optimality concepts and system properties; on the existence, characterization and computation of optimal policies; on the development of useful heuristics with assured performance; and on identifying and exploiting the structure of specific applications to gain insight about them.

Lattice Programming 

This area is concerned with predicting the direction of change in global optima and equilibria resulting from changing conditions based on problem structure alone without data gathering or computation. Rooted in the theory of lattices, this work is also useful for characterizing the form of optimal and equilibrium policies, improving the efficiency of computation and suggesting desirable properties of heuristics. Applications range widely over dynamic programming, statistical decisions, cooperative and noncooperative games, economics, network flows, Leontief substitution systems, production and inventory management, project planning, scheduling, marketing, reliability and maintenance, etc. Recent applications include price and warranty setting in the automotive industry, and optimally stepping up pressure in gas pipelines.

Network models are widely used in industry, government and engineering for supply, distribution, manufacturing, transportation, communications, construction, mining, investment, scheduling, sequencing, routing and reliability. Networks and graphs also serve as fundamental tools to study the structure of matrices, Markov chains, probabilistic dependence, optimization problems, etc. The research in this area focuses on single-commodity, multi-commodity, dynamic, equilibrium and stochastic network flow and design problems. The costs typically exhibit economies, diseconomies or constant returns to scale. The emphasis is on the identification and/or development of the relevant structural properties of such systems; efficient methods of finding optimal or near-optimal flows, designs and equilibria; and on applications to a wide variety of industrial, public and engineering problems.

This area focuses on the development and analysis of models to facilitate efficient management of inventories of products and service capacity. These problems are addressed in environments in which there is often uncertainty about demand, supply, prices, quality and product life; costs exhibit economies, diseconomies or constant returns to scale; there is often competition; and decisions are made sequentially as new information is acquired. Issues addressed include: where, when and how much to stock at various points of a supply chain; how to price products and services; how to size and time expansion of capacity; how to respond to competition. Recent applications include optimal overbooking policies for airline seat inventories and optimal paper-mill supply policy.