97 Things Every Engineering Manager Should Know

Chapter 51. Manage Complexity with Diversity

Mike Fisher

As engineering managers, we often develop and manage complex systems that have interacting components (people, process, technology), the combination of which make it very difficult to predict the outcome from simply observing the inputs. Additionally, many of the subsystems are tightly coupled in that a failure of one can result in cascading failures. As such, managing the inherent risk associated with these complex systems is a daunting task.

Decades of research into the management of complex systems has given us normal accident theory, high reliability, and the logic of failure, just to name a few. Organizational design based on high reliability even identifies five characteristics of organizations that are best to deal with complexity: preoccupation with failure, reluctance to simplify interpretations, sensitivity to unexpected conditions, commitment to resilience, and deference to experts. The logic of failure identifies ways to fail in the decision-making process, including failing to prioritize goals, using overly reductive models, and failing to generate alternative plans. However, only recently have researchers turned their focus to the makeup of the organization instead of its behavioral characteristics and processes.

Research has begun to catch up with what savvy practitioners have long known, that diverse ...