A critical part of a proposal for the development of a complex system is its cost. One important element of this cost is the effort to perform the systems engineering function. The prevalent method for estimating these costs is a "bottom-up" engineering estimate based on the experience and judgment of the estimator. In some cases, analogy to actual costs of previous, Similar programs is employed. Neither of these methods provide the accuracy, consistency, or defensibility needed.

Various methods and models are used to assist in estimating costs in other disciplines across differing applications. These methods were evaluated for suitability in estimating systems engineering costs for a computer based electronic defense system. The requirements and conditions, both technical and organizational, for developing a cost model are delineated. Considerations for automating the model are discussed.

Major findings are: • Models used in conjunction with traditional estimation methods can improve the overall systems engineering cost estimation process,

• Activity based parametric and analogy models show the most promise in the near term due to the reasonable accuracies achievable, the high speed and low cost with which estimates are generated, and their adaptability to automation, and

• Before a model can be developed and be a useful tool within an organization a number of conditions must exist, primary among which are the adoption of a standard cost structure for bidding and managing systems engineering efforts, the creation and maintenance of a cost/parameter database to allow entry and analysis of actual program cost data, and management acceptance and commitment of adequate resources.