Complex Adaptive Systems
Safety campaigns, safety meetings and demands for better work, although well-intentioned, often do little to nothing in influencing how a system operates. They offer no practical solutions by those unaware of the complex system dynamics.
Complex adaptive systems are networks in which individual and collective behaviors dynamically adapt and self-organize in response to changes. They run at maximum capacity, continuously evolving with technological advancements. This dynamism often positions them at the brink of failure, yet these systems seldom collapse.
The behavior of these systems is both unpredictable and resilient. They adapt to various pressures, allowing operators to monitor, seize opportunities, estimate proximity to failure, and learn from system behaviors.
Economic pressure pushes these systems away from economic failure, while workload management aims to reduce stress. This constant flux can drift the system's operating point towards a potential failure boundary. Implementing rules or reacting to incidents readjusts the operating point, influencing system stability and safety.
Accidents or major failures push systems beyond normal limits, demanding urgent action and nudging the system against economic boundaries. These incidents rarely cause lasting shifts in the system's position, as economic pressures persistently push the boundary. The margin of safety acts as a warning zone. Despite its presence, systems still experience failures. The challenge lies in understanding this margin relative to the elusive accident boundary, often unknown until a catastrophe occurs. Over time, repeated successful operations sometimes desensitize operators to risks, leading to a normalization of deviance. This gradual shift of the margin boundary towards the accident boundary moves the system's operating point closer to potential failure.
Despite these challenges, systems remarkably continue to operate. Understanding this phenomenon lies in short-term monitoring, reaction, anticipation, and learning. So, in complex adaptive systems, continuous adjustments and a deeper understanding are needed to ensure stability and reliability.
Illustration: Rasmussen, J. (1997), Risk management in a dynamic society: a modelling problem, in: Safety Science, Volume 27, Issues 2–3, November–December 1997, Pages 183-213.