Optimizing DTC in case-based development with parametric modeling tools

Abstract

In Internet of Things (IoT) world, there can be many sensors deployed in a system, thus their costs must be considered during the system design and material selections, in order to support massive production and deployment. Design-To-Cost (DTC) is an effective technique for this situation. In Model-Based Design (MBD), DTC cost estimates are preferred to be generated from models as one of the parameters to optimize the design. However, most cost modeling methods focus in estimation on costs associated with new product development, and much fewer discuss the estimate of the switching costs for new products that are designed leveraging previous products. In this paper, we explore certain available tools for DTC for such purposes and discuss their advantages/disadvantages. We performed an experimental study on a building PHM (Prognostic and Health Monitoring) sensing system design example, and demonstrated an approach to achieve switching cost estimation using parametric modeling by treating part of the new product as another prior product. Finally, a multi-objective optimization is used for design selection, to achieve the best trade-off between cost and performance.