Distributed, integrated model of DOE weapons complex to hone efficiency with information technology

The weapons complex is turning to information technology to hasten a revolution in engineering. This approach promises significant gains in efficiency, says Len Napolitano, Manager of Advanced Concepts Dept. 8130. He heads a project that is creating a distributed, integrated model of the DOE weapons complex. The model will be used to help predict the consequences of decisions about issues ranging from dismantlement requirements to refurbishment schedules to capital investments, helping decision-makers evaluate whether the complex can meet changing demands with anticipated resources over the next 10 to 30 years.

The project — intended to get the right information to the right people at the right time — incorporates existing and new models of different aspects of the complex.

By piecing together sub-models, "a high-level question about the impact of START III can receive a more realistic answer," adds Michael Johnson (8114), who is principal investigator of the project, called CoMPASS — a Confederation of Models to Perform Assessments of Stockpile Stewardship. CoMPASS has been under way nearly a year, already receiving positive feedback.

Astonished at integration, links

"People have been very astonished at how we’ve been able to integrate information and provide links that they didn’t think were there," Len says. "A lot of projects have the goal of reducing cost and time by a factor of two. One way we can do that is by using information technology."

Len compares the challenge to using a map to reach a destination. Ideally, the map will provide details about only the best path, reasonable alternate routes, and important landmarks without other unnecessary information. Decision-making then becomes less of an exploration, just as using the Web is easier with a search engine. In fact, the model uses a browser to lead to appropriate answers. Woven together, the disparate models of aspects of the weapons complex thus become much more powerful than the sum of these pieces. Once the enterprise is modeled in a static snapshot that illustrates products from the viewpoint of components, plant capabilities, equipment, and staff, it is then possible to simulate over time how these factors would change under dismantlement treaties or parts-replacement programs. Business practices can then be fine-tuned in anticipation of changing needs.

The promise of enterprise modeling is being pursued against a backdrop of fixed resources and changing priorities, adds Michael. Weapons are aging past their originally intended lifespans. The complex is shrinking. Experienced staff are retiring. Various alternative activities within the stockpile life extension program will compete for the same staff and facilities. In the "lean" complex that now exists, workload resources and capacity must be carefully balanced across the complex now and into the future.

"It’s a really complicated problem," Michael says, "with a lot of parameters."

The project is the outgrowth of work that began four years ago, as the fate of the B83 stockpile was being computer-modeled by Sandia. Factors such as an aging population, limited critical resources, and established policies and procedures were similar to those of another enterprise, a health maintenance organization. So, Sandia began working with Kaiser Permanente under a funds-in cooperative research and development agreement to understand return on investment of capital decisions, the effects of changes in health care practices, and the use of time-evolving patient models (or, for the B83, time-sensitive reliability models).

Ten sub-models being integrated

This fall, Michael is demonstrating integration of 10 sub-models from sites across the country into a common framework. The framework has to interface with commercial, off-the-shelf software packages that support seven different applications, gluing them together in a secure, seamless fashion.

Sandia is taking the lead on enterprise modeling in this effort, drawing on its role as a system integrator for the complex. The trend toward enterprise modeling is occurring in other industries, Michael points out, as businesses also discover the utility of interrogating rich databases of documents and facts with simple browser-like pointers.

Here, the process will use Sandia’s Product Realization Environment tools that enable users to easily access and manipulate large sets of data, analytical methods, or engineering design packages in a nearly transparent fashion. The models are being linked in the weapon laboratories’ secure network, where different levels of access are controlled through a single gateway.

ADAPT initiative funds project

The project is funded under the Advanced Design and Production Technologies (ADAPT) Initiative, as part of the Enterprise Integration Program. This program focuses on providing the secure, integrated information infrastructure and the enterprise software tools needed throughout the weapons complex to support product design, development, and production activities for stockpile life extension.

Eventually CoMPASS will fold together dozens of models at seven major facilities, addressing such diverse aspects as production, planning and environmental regulations, stockpile surveillance, and transportation.

The CoMPASS team includes Heidi Ammerlahn, Ann Yoshimura (both 8112), Troy Delano, Kristen Nostrand, Tim Sa (all 8114), and Todd Plantenga (8950).