To keep pace with the design challenges of the modern day, traditional design approaches are no longer sufficient. Computational models have been playing an increasing role in the design process, but new methods are needed to manage complexity and to understand the effect of uncertainties, which could stem from many sources – technical, economic, regulatory, and organisational.
In this project, IDC researchers are developing new methods to address these challenges, with particular attention to the role of complexity and uncertainty in the design process. The research draws on concepts from estimation theory, control, and information theory. We move away from the traditional approach that attempts to pose the design task as an optimisation problem, and instead view the system design as an estimation process. This estimation viewpoint leads naturally to methods to manage uncertainty: quantification and propagation of uncertainties to estimate uncertainty in output quantities of interest (such as system performance, cost or requirements satisfaction), formal tracking of uncertainty through the design process, and sensitivity analysis to identify important sources of uncertainty and to guide resource allocation.
We are currently developing the methodology and demonstrating it on simple circuit examples, with a goal of transitioning to more complex problems relevant to the IDC grand challenge applications.