recommendations made by the committee in different disciplinary areas, identifies common themes, and also provides recommendations on bridging design and manufacturing in DoD acquisition as well as in engineering education.

SYSTEMS ENGINEERING

Recommendation 1. Systems Engineering: The Department of Defense should develop tools to facilitate the definition of high-level mission requirements and systems-level decision making.

Tools to create, visualize, and analyze design and manufacturing alternatives can facilitate systems-level decision making. A specific opportunity is to develop tools for converting customer needs into engineering specifications, and for decomposing and distributing those specifications to subsystems and components.

The design and manufacturing process leading to product realization is essentially a system of systems. Performance requirements are set at the highest level. For example, in an automotive system, the vehicle capacity, performance, weight, and cost are specified. In weapons systems, range, power, and cost are specified. These requirements come from analysis of the needs of the customer and from estimates of funding and other resources.

Performance requirements, set at the highest level, flow down to the other levels in the form of system and interoperability specifications. Conceptual designs are broken down into subsystem and component designs. Decisions are then made about materials, assembly, and manufacturing processes. Information may also flow back up this chain to modify the design.

Such a sequential approach, however, can lead to inefficiencies. Decisions may be made at one level without full consideration of the implications for other levels. For example, parts may be designed that cannot be manufactured or parts can be manufactured that are difficult to assemble. Simple manufacturing processes may be impossible to use because of an arbitrary design specification. A systems engineering approach can avoid these consequences by requiring collaboration at different levels and collective decision making.

Moving from a linear approach to an integrated systems-level approach will require substantial cultural and organizational changes. In order for such an approach to work, all of the participants require access to sufficient and timely information. Designers need to be able to work with a multidimensional trade space, where design alternatives can be effectively compared. Software tools can make important contributions to this effort. The systems approach requires that analyses and decisions be made in multiple disciplines and that global optimization be performed. Further, since it is likely that several different analyses will be done, functional interoperability between the various computer codes is essential. Finally, the integration of such tools will require significant training.

Research is needed on improved techniques to derive and elicit mission needs, and to translate those needs into model-based requirements and executable specifications. The current process for this step is largely experiential and tends to force designers to think within the constraints of previously developed products, discouraging innovation in favor of incremental improvement.

So that designers can link system requirements and full component specifications, better tools and techniques are needed to create, visualize, and analyze the design trade space. These tools will include fully interoperable codes, where data do not have to be recreated for different analyses. They will also support automated abstraction of the data at different levels.

The DoD can foster the growth of this systems approach in two ways. Funding can be provided for demonstrations and benchmarks of existing tools. This funding would also encourage improved interoperability. The DoD can also support systems engineering curricula



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement