Thursday, June 3, 2010

SDM thesis addresses obsolescence mitigation in complex systems - SDM Pulse Summer 2010

By Jaime Devereaux, SDM ’08

Jaime Devereaux SDM ’08
Editor’s note: In this article, Jaime Devereaux, SDM ’08, outlines the major points covered in her SDM master’s thesis, “Obsolescence: A Systems Engineering and Management Approach for Complex Systems.” Devereaux is a manager in systems engineering at Raytheon Integrated Defense Systems.

Too often, obsolescence mitigation is only considered once obsolescence has become imminent. But such mitigation is an increasingly important aspect of large systems development and maintenance because the life cycles of components are often up to 10 times shorter than the life cycle of the overall system.
Currently, recommended system-level obsolescence mitigation practices typically exist for the early design phase of new systems. Obsolescence mitigation slows the onset of obsolescence and makes systems flexible enough to change as necessary when obsolescence looms. Mitigation techniques include use of open architectures, standard interfaces, model-based architecture, and advance planning—that is, designing the system with potential future requirements in mind. Unfortunately, many large, complex legacy systems were rarely adequately designed for obsolescence mitigation as these techniques were not commonly used. For some systems, a choice is made not to use the design approaches above due to pressures related to the initial design cost and schedule or because the system is intended to be replaced prior to the onset of obsolescence. Many systems are in use longer than the life cycle of their components, due to the excessive cost and time needed to design a replacement system. In these cases, a different approach is necessary.
There are many different types of obsolescence. Psychological obsolescence, for example, drives the consumption of products that rely heavily on styling (clothes, cars, etc.); these products can become “old” before the end of their physical life. Quality obsolescence drives the consumption of disposable goods such as razors and plastic silverware, which are designed for much shorter lifespans than their more robust counterparts. In my research, I focused on the role of technical obsolescence and manufacturing/maintenance obsolescence on large, complex systems.
Technical or functional obsolescence occurs when a new product enters the market that performs better than anything that was previously available. Manufacturing or maintenance obsolescence occurs when there is no need in the current application for a product with increased function—and market demands do not support a supplier’s continued production or support of the older component.
Possible causes of obsolescence in large, complex systems include:
•    Use of commercial, off-the-shelf technology—while using such components reduces the kit costs associated with purchasing the component, outside components increase the number of influences on a system through vendor supply and support contracts, which may increase life-cycle costs
•    Increased use of electronics in modern systems, which shortens the life of many components
•    Market changes that make a particular component unavailable
•    Prohibitive costs for continuing to manufacture components based on older technology
•    Changes in the system’s environment that make the original function obsolete or suboptimal, requiring evolution or replacement of the system to meet new needs
•    Corporate management decisions that influence how much human intellectual capital can be leveraged to change a system long after its initial design
Reactive obsolescence mitigation techniques tend to focus on identifying obsolescence in the hardware through supply chain monitoring, classifying the obsolescence, and developing replacement solutions. But these efforts don’t fully acknowledge the impact of these choices on the rest of the project or system. On the whole, obsolescence mitigation approaches have not made use of the engineering change analysis techniques taught in MIT’s System Design and Management Program (SDM), including the design structure matrix (DSM) and change propagation analysis using the change propagation index (CPI). CPI in particular is useful for evaluating how components within a system propagate changes to other components in response to an external change. 
Figure 1: This design structure matrix shows the system-level
baseline for the weapon system Jaime Devereaux, SDM ’08,
considered in the case study. Note that the specific row and column
descriptions have been deleted to protect proprietary information.

In my SDM research, I attempted to create a systems-level, full-life-cycle mindset for determining a system’s optimal obsolescence mitigation strategy. By combining recommended approaches for obsolescence mitigation gained from a literature review with the experience I gained interviewing key experts for a real-world case study (see Figure 1)—and by incorporating SDM systems engineering techniques for dealing with change into my analysis—I was able to zero in on a more robust systems engineering and management approach for dealing with obsolescence. 
Figure 2. The above chart illustrates a systems engineering
and management approach for obsolescence mitigation
The proposed approach, shown in Figure 2 and Table 1, allows for mitigating obsolescence in a large, complex system in both a reactive and proactive manner. Step 1 requires the engineer to understand the system-level architecture before evaluating the impact of changing a component of that system. Figure 1 illustrates a DSM of the case study used in my research (a weapon system). The nth row and nth column have the same variable, in this case subsystem, under consideration. The matrix then shows how the element in the nth row and mth column are related using the legend table. Further, the specific row and column descriptions have been deleted to ensure the proprietary nature of the work can be respected. Once the system-level relationships and interfaces are understood, the relationships can be drilled down to the subsystem level and below if necessary. In addition, in Step 2 these DSMs can be modified to help calculate how likely it is that a change to a given component will impact other system components. The checklist is designed to help the systems engineer consider the many impacts that obsolete components can have on the overall system or program.
Table 1. Checklist for obsolescence mitigation
As systems get more complex and the life cycles of components decrease, it is essential for companies to acknowledge the complexities surrounding designing for and adapting to obsolescence. The intent of this framework is to give engineers and program managers a starting point for inquiry and an approach for evaluating the impact an engineering change due to obsolescence may have on the system as a whole. By looking for these impacts before change is imminent, engineers can ensure that changes to the system are dealt with proactively rather than reactively. In addition, changes will likely be identified earlier in the engineering change life cycle, reducing possible schedule and budget impacts from “surprise” discoveries.
For more information on the obsolescence mitigation approach described here and for references that can help you develop an obsolescence mitigation strategy of your own, please write for a copy of my thesis, “Obsolescence: A Systems Engineering and Management Approach for Complex Systems.”

No comments:

Post a Comment