 | |
Multi-agency group could improve the effectiveness of its strategic budgeting using a scenario-based optimization model
| |
 |
Business impact Improved understanding of the costs and benefits of fire preparedness alternatives will result in increased national effectiveness.
Issue The Office of Management and Budget (OMB) wanted a system that was consistent among all federal wildland fire fighting agencies, so that it could produce more comprehensive budgets for fire management.
Executive summary To improve the results of a budgeting process done two years in advance, the five federal wildland fire fighting agencies are developing a scenario-based budgeting system to conduct cost effectiveness analysis.
What IBM did These agencies were using three different processes, two different budgeting systems and individually customized models to develop their budgets. The agencies wanted to build an application that would support a common inter-agency process for fire management planning and budgeting. The new application would also need to produce more effective budgets for a planning process that is done more than two years in advance of an actual firefighting season.
Wildland fire factors (such as weather, ignition sources and topography) are impossible to predict two years in advance. This level of uncertainty in budget planning required using a scenario modeling and optimization approach rather than traditional simulation forecasting methods. IBM Research is applying world class skills in optimization and modeling to develop this solution.
Initial improvement of the budgeting model was begun by government and university researchers. Working with these researchers, IBM Research developed a model that considers various factors that affect fire growth and containment. Optimization techniques allow the system to identify the most effective initial attack organizations for a range of budget levels.
As the budgeting model and framework were developed, IBM consultants began working on the architecture and application that will be deployed in September 2004 for the 2007 budgeting cycle. A user-friendly, Web-based interface will facilitate input of key variables that describe the probability and behavior of wildfires. Over time, interactive maps will be added to the system making budget allocation faster and easier to understand. All in all, ease of use and a common framework are expected to produce more comprehensive budgets as well as facilitate inter agency collaboration.
Capabilities applied IBM combined Research's experience in scenario-based modeling with the IBM Global Business Services consultants' experience in enterprise architecture and application development. Working together, they are developing a system that will work not only for preparedness, but will also serve as a foundation for future fire management budgeting initiatives.
| |
|
|
Related case studies | 
|
More on the team | Gyana Parija For optimization experts like Gyana Parija, stochastic optimization is all about using scenario trees to support decision-making in uncertain environments. |
More on research | IBM Research: Algorithms & Theory Learn about IBM Research's distinguished history and involvement in the development of algorithms and the theory of computation in a variety of areas. |
More on research | Stochastic Programming at IBM Research Learn about IBM Research's work and involvement in the area of Stochastic Programming. |
More on research | Large Scale Strategic Budgeting for Wildfire Management This research develops a strategic planning model for developing preparedness budgets necessary for managing wildfires during the initial response period. |
|
