Jets, locomotives, people, trucks,energy, vaccines, information, money... The management of physical, financial and informational resources is the domain of CASTLE Lab. We want to know how to use resources efficiently in the presence of different forms of uncertainty. We also want to know what information we should have to make these decisions, and how much we are willing to pay for this information.

We solve these problems using a relatively new class of algorithmic strategies known as approximate dynamic programming. This work has produced a merger of dynamic programming (Markov decision processes), math programming (linear, nonlinear and integer), simulation and statistics. 20 years of research in this area has produced a new book, Approximate Dynamic Programming, published by John Wiley and Sons.

In a recent project, a model based on approximate dynamic programming was shown to closely simulate the dispatching decisions of Schneider National, one of the nation's largest truckload motor carrier with a fleet of over 15,000 drivers. Modeling this process required solving a dynamic program which featured a state variable with millions of dimensions. In addition, the model had to handle the complex operational details required to properly model the management of drivers and loads. For a paper on the project to appear in Transportation Science, click here.

These modeling techniques have been used in a number of other products supported by the industrial sponsors of CASTLE Lab. CASTLE Lab could not exist without the active participation of the sponsoring companies who do much more than provide financial support. They provide data, feedback and an opportunity to see if the ideas work in the field.

To download a high-level summary of CASTLE Lab highlights, "From the laboratory, to the real world, and back" click here.

I hope you find the material interesting, and perhaps useful. If you have any questions, please contact me.

Warren B. Powell

powell@princeton.edu

(c) Warren B. Powell, 1997-2008