Modeling and Simulation

Group W people design and create large-, medium- and small-scale simulations to help analysts, researchers and clients explore and understand complex problems. When employed appropriately by skilled analysts, simulations are invaluable tools for exposing elements of a problem and evaluating alternative solutions. Misused, they can obscure true solutions and compound the difficulty of understanding a problem. Visionary clients know to carefully scrutinize a simulation and its application when weighing results.

The design of a simulation starts with a conceptual model—an abstract, algorithmic representation of the various entities and interactions of interest. It takes exceptional people to get this right—a poorly conceived model cannot be redeemed by excellent programming. Further insights into the problem and how to attack it are often-overlooked benefits of this stage. Our model designers look to uncover previously hidden issues and illuminate their implications for the analytical objective.

... I believe the best test of a model is how well the modeler can answer the questions what do you know now that you did not know before? And how can you find out if it is true?

- Jim Bower

Our team of model designers, software architects and computer scientists turn a conceptual model into a functional simulation. The challenge in this stage is to enforce adherence to the design while promoting innovation in the implementation. As computer programs, simulations can involve hundreds of thousands or even millions of lines of code. They typically continue to evolve through a series of upgrades and can have useful lifetimes of more than a decade.

The creation of a useful, elegant and durable piece of software is an exercise in striking a balance between often competing objectives (e.g., fidelity, runtime, data availability, transparency). Building a useful and usable simulation also demands a careful combination of science and art. Scientific principles guide the design and implementation, but art and wisdom must be applied to ensure that real analytic resolving power is the result.