Resilience Engineering #7: Dynamics of Success – Functional Resonance Accident Model

by Gary Monti on July 19, 2011

The Holy Grail for complex organizations experiencing high risk is finding a balance between stability and flexibility. This presents a very real challenge since the environment is almost always shifting and the team has to think on its feet because time, money, people, and other resources are limited. There isn’t enough time to cycle up to senior management and back down to the team.

The previous two blogs presented linear models of success and failure that are inadequate in complex situations but which are still alive and well in many organizations. They are also limited in term of being either fixed solely on the individual (Domino model) or top-down in terms of policies and procedures (Barrier model).

This blog starts the process of looking at a more realistic model for addressing success and failure in dynamic situations, the Functional Resonance Accident Model (FRAM) developed by Hollnagel. Its roots are in complexity theory and it comprises four principles:

  • Equivalence of success and failure. Successful teams rely heavily on anticipatory awareness, i.e., paying close attention to the environment as it is, without expectations. They perform early-warning weak signal analysis, and decide how best to organize for the situation. An anesthesia team might best characterize this behavior. Guiding medical principles are present but the number of hard-and-fast rules is low compared to how much the anesthesia team must monitor the surgery and think on their feet constantly assessing the entire situation while simultaneously monitoring details. Failure can occur when the team temporarily losses this ability.
  • Approximate adjustment. The team is constantly adjusting its performance to suit the situation. This includes adapting to shifts in resources as well as unique requirements for the specific task at hand. Imagine your elderly, sick grandmother is staying with you and she is very sensitive to excess heat but also chills easily. You have an air conditioner that can maintain 75°F indoors in direct sunlight only if the outside temperature is below 95°F.  On days forecast to be hotter than 95°F what do you do? You must gauge what time in the morning to turn the thermostat below 75°F. How low do you turn down the temperature? At what time do you do it? Does it vary with the afternoon forecast? Could she chill with the setting you’ve chosen? Answering these questions from day to day is making an approximate adjustment in the presence of limited resources and high risk.
  • Emergence. The constant adjustments in performance means there is constant variability. This variability can have a compounding effect, which is non-linear and disproportionately large. New behaviors can emerge. A tipping point can be reached. Think of the impact one failed safety relay has had on the electrical grid in the United States. Whole areas have been plunged into darkness.
  • Functional resonance. A whole constellation of variables can show emergent behavior and impact each other, causing a particular function in a system to resonate without there being one direct, cause-and-effect relationship to which one can point. Think of the speed with which Google grew initially or sales of the iPad or the initial impact of Palm. Failure can emerge as well. Think of Palm’s sales for the last few years before being bought by HP. In a different area, look at how the functional resonance of political dissent has changed in the Middle East. Have changes in communications had an impact?

In principle you can see that FRAM is much more robust than the Domino or Barrier models covered in previous blogs. It goes well beyond the individual or attempts to create all-encompassing policies and procedures. It addresses the dynamics of the situation, which keeps it grounded. We will go deeper into the FRAM model in the next blog.

Related Articles

Previous post:

Next post: