Discovering Failure Modes Early in the Design Process

Earlier this month, I spoke with Ed Henshall, who just recently published a book entitled Right By Design: A Novel Approach to Failure Mode Avoidance. His book presents an approach to product design based on Failure Mode Avoidance that utilizes a series of strongly interrelated engineering tools and interpersonal skills that can be used to discover failure modes early in the design process. The tools can be used across engineering disciplines.

During our conversation, I asked him: "Is it possible to discover failure modes early in the design process?” Here is his complete answer:

The short answer -- Yes, with difficulty. 

In looking at a longer answer, I would rephrase the question slightly --  “Is it possible to discover potential failure modes early in the design process before you have a design?”  

Firstly, the word “potential” is important as it indicates that the design can fail but has not yet failed. Secondly, by not having a design I intend that the design is fluid and not finalized meaning that it can readily be changed without impacting the cost and timing of the design process. This latter point is the good news -- if failure modes are found that require fixing early in the design process, this can be done inexpensively. However, the associated bad news is that it is difficult to discover failure modes early in the design process when a design is fluid. 

The key to this conundrum is to have a clear understanding of what it is that the design is intended to do, and its function, along with an equally clear understanding of the way in which the design will achieve this function. What is important here is that the design is initially considered from a functional perspective rather than a hardware perspective. To quote the well know architect Louis Sullivan, “Form ever follows function.”

The System State Flow Diagram provides a way of modeling a design from the functional perspective allowing potential failures of function to be identified in a rigorous and systematic manner. This enables design countermeasures subsequently to be developed in moving into hardware design. 

Discovering failure modes early in the design process requires effective and efficient teamwork, which does not happen as a matter of course when groups of people work together but requires significant attention to, and coaching of, the team process.  

What do you think of Ed Henshall's perspective? Does your organization have effective "team process" and leadership?