Abstract

In aerospace and in many other fields of engineering, it is common practice to forecast the behavior of a physical system by analyzing a mathematical model of it. If the model is accurate and the analysis is mathematically sound, forecasting from the model enables an engineer to preview the effect of a design on the physical behavior of the product. Accurate mathematical forecasting reduces the risk of building latent design errors into the physical product. Preventing latent design errors is an important part of successful engineering. If a product contains a latent design error, it can cause operational malfunction. When a latent design error is detected, removing it requires backtracking in the product development cycle. This backtracking can consume large amounts of time, money and human resources. Although digital computers now are embedded as operational components in many aerospace and other physical systems, capabilities for mathematically forecasting the physical behavior of computer programs are only now beginning to emerge. Without these capabilities, latent design errors in computer programs frequently go undetected until late in program design or until the program is tested or even until it is in actual operation. This increases the dual risks of operational malfunction and high resource consumption caused by developmental backtracking. When computers are embedded in other physical systems, these

Citations