Models for platform-based product family design

Abstract

Platform-based product families have been an effective way for companies to offer increased variety into markets, while containing the resulting complexity of developing large numbers of products. A product platform is the set of resources - components, processes, technologies, and knowledge - that are shared across multiple products offered by a firm. The products derived from that common platform are called the variants, and the set of variants forms a product family. This thesis presents a theoretical basis for modeling the design of platform-based product families, as well as practical implementations based on those models. The problem is formulated as an optimization, where the requirements of the desired variants must be balanced against family objectives, such as maximizing the value of developing the family. A general case is formulated that covers the effects of using a platform-based design on the benefits and investments required to produce the desired family. The problem is then simplified into a two-step optimization approach to apply it to actual design situations for complex products. The first stage considers the technical details of creating feasible product families that satisfy the variants' requirements while optimizing the expected value to the firm. The second stage evaluates the design alternatives generated by the first step, considering the effects of uncertainty during the actual development of the family on its value to the firm. This evaluation is then used to select the most appropriate choice of family design. A case study of the design of multiple NASA exploratory space missions based on alternative telecommunications technology platforms is shown. Applying the approach resulted in several alternative family designs, some of which had not been previously considered viable. The resulting candidate designs were evaluated through the use of decision analysis models developed in this thesis, which calculate the value to the firm of each design. Simulation was then used to evaluate the value of each alternative and its flexibility to changes in uncertain factors during the development process. As a reference, a Pugh-type selection method (a qualitative, multi-criteria approach) was also explored and applied to the example of platform-based spacecraft design.