Editor’s note: Augustin Friedel is a master’s student in mechanical engineering and management at Germany’s Technische Universität München (TUM.)
For the last six months, I’ve worked on a research project investigating uncertainties in product platforms at MIT’s Systems Engineering Advancement Research Initiative (SEAri). The research was guided by SEAri principal research scientist Dr. Donna H. Rhodes, and TUM Aerospace Professor Eduard Igenbergs.
I saw the huge potential of the platform approach in complex systems while working on a platform project for a client of a consulting company that specializes in automotive, transportation, and aerospace systems engineering projects. I also became aware that the process of platform design was, in general, not explicitly understood among members of teams working on a common problem.
I realized that some of the confusion and uncertainty around the platform development process could be reduced through structured methods. This would allow teams to develop successful product platforms more efficiently and turn their companies into market winners. Complex systems, including product platforms, are subject to uncertainties that may lead to suboptimal functional performance or even catastrophic failures if unmanaged over time. Identifying uncertainties in the front-end—and implementing ways to mitigate problems that may occur—can be a part of the product platform design process that adds value to the platform as a system.
My thesis describes the journey of developing a method for investigating the management of uncertainty that will help platform developers with the complicated process of transforming an idea into a finished product. In it, a platform is defined as a set of architecture, common modules, and interfaces from which a stream of derivative products can be efficiently developed and launched.
The architecture is the configuration within the product; a module is a part or a group that allocates a function to the product; and the interfaces are connections between the modules and architecture, among the modules themselves, and between the platform and customized parts of the product.
The thesis has three sections. In the first, I discuss the results of an empirical case with 10 participants (on hierarchy levels between lead engineer and senior project director in automotive, electronics, agriculture machinery, and defense industries), studying the uncertainties in product platforms and the consequences. This study found that most uncertainty occurs for reasons that could have been predicted. Furthermore, companies often realize too late that there is a problem because of an uncertainty, reacting only after discovering that the platform didn’t meet expected performance criteria.
Given the findings, there appeared to be a need for an approach to managing uncertainty that would allow a value-robust platform to be built. In the second part of the thesis, I describe developing a framework for that purpose.
The preliminary framework has seven steps, providing a tool for managing uncertainties and risks within a platform (see Figure 1). The first two steps are based on the Quality Function Deployment method of transforming customer requirements into engineering metrics. The third step is based on the ISO standards for managing risk by identifying, analyzing, and evaluating different uncertainties. The fifth step describes an approach for treating the uncertainties and risks of implementing mitigation mechanisms in critical parts of the platform.
|Figure 1. Overview of the preliminary framework for managing uncertainties in product platform lifecycles.|
Critical parts are identified on two different paths in Step 4 of the framework. The first starts with tracing the impact of uncertainties on the components; the second path traces the impact of each uncertainty on the engineering metrics. (For predicting an instance where exercise of a mechanism is needed, the Epoch-Era Analysis developed by SEAri Research Scientist Adam Ross and Rhodes was adapted.) Step 6 compares platform designs and selects the most valuable one with a Multi-Attribute Tradespace Exploration methodology, developed by Ross. Step 7 runs parallel to the other steps, and the monitoring aspect allows reaction on uncertainty before the risk can take effect.
The third section of the thesis successfully applies the framework to the example of a platform-based cleaning robot.
Future research can serve to evolve and further test this framework. The final thesis, “Investigating the Management of Uncertainty in Product Platform Lifecycles,” as well as more information on SEAri and its research, is available at seari.mit.edu.