V&V&UQ in Engineering

Engineers rely on quantitative models to help guide their design, production, and operational decisions. The success or failure of an engineering project often depends on how well the engineering analysts involved handle their model predictionsFor generations, engineers have done what they can with best-guess analysis and factors of safety. For the most part, they have done amazingly well. Nevertheless, our civilization keeps pushing the boundaries of technical possibility with successively larger and more daring engineering projects, and the best-guess approach to engineering analysis has faltered many times. On a good day, only money and valuable equipment are lost. On a bad day, lives are lost. Whether a model is improperly executed mathematically, or the model itself is just plain wrong, or uncertainties in the model parameters are unrepresented or even just improperly represented; poorly managed predictions can lead to bad engineering decisions, which can lead to disaster. The purpose of V&V&UQ is to put engineering analysis on a more solid footing, by systematically scrutinizing model predictions for their mathematical and empirical rigor.

Alexandria Validation Consulting, LLC exists (in part) to help make a new generation of large-scale engineering projects possible. We do this by providing the V&V&UQ tools necessary to support safety analysis on those systems. That being said, our interests and methods are not limited to system-level analysis. Large-scale systems consist of successively smaller components. Every problem solved leads to new insights and new approaches. We are interested in any engineering problem of any scale in which traditional risk analysis methods have failed to yield a sensible and trustworthy answer.