Comparison on How Do We Fix Systems Engineering and Learning from Failure

1. Comparison of Perspectives on Process

When considering the perspectives of both How Do We Fix Systems Engineering (HDWFSE) and Learning From Failure (LFF) with respect to the role of process in systems engineering, Griffin sets a tone of healthy respect when he says, “…this is no small feat, and the methods, processes, and tools which have developed over the last half-century to formalize and systematize it as an essential engineering discipline are not to be slighted.”^[1]  These work flow concepts are inextricably linked together within the discipline. Yet, very quickly he states that when one considers the state of the art, new systems engineering processes are likely not the answer in response to failure of a system.^[2] 

The panel discussion that created LFF takes a similar tone in many respects as HDWFSE which is not a surprise since Griffin is on the panel. Early in this article, the authors present a deeper insight into why it is counter-productive to pile more processes into a product creation cycle: Namely, that organizations can become risk averse which in turn leads to a declining spirit of innovation. An abundance of engineering processes and controls can shackle organizations from being able to meet strategic objectives.^[3] 

Where Griffins paper laid the ground-work in asking about fixing systems engineering, the panel paper provides an inquiry-based approach to find answers on how to deal with failure through the proper reaction instead of additional processes.^[4]  Interestingly, the panel does not provide specific concepts, in contrast, Griffin’s HDWFSE paper points to the principle of elegance to combat unintended emergent properties.  

The LFF panel recognized that adding process is a natural reaction, even an addiction, however, most members of the panel claimed that failing is essential to learning, and that by extension success most often denies the opportunity to learn.  In the case of the X-33 reusable launch vehicle, the chief systems engineer set the tone that if teams weren’t failing that meant that they weren’t stretching technological development far enough.^[5] 

In LFF, the panel succinctly summed up their position with Kadish’s point that the reason more process is not a solution for dealing with failure is that process cannot replace leadership or understanding and a good process well-formed does not prevent failure.^[6]

1. The Impact of Interfaces and Information Sharing in the SEBoK Case Study

The impact of interfaces and a lack of information sharing between NASA and ESA primarily resulted in creating poorly understood single points of failure, the joint effort of both agencies and their cadre of suppliers created an additional level of people-systems interface that neither team was prepared to deal with appropriately.  This political situation and the choice of solution boundaries (one agency takes the lander and one takes the orbiter) produced and exacerbated the failure modes of the separating functions.  The situation was completely missed with the cancelling of full system tests before launch due to budget constraints.^[7]