Skill atrophy

Skill atrophy is the degradation of human operational competence that occurs when a task is performed by an automated system instead of by the human operator. Unlike simple disuse, skill atrophy in automated environments is systematic: the automation removes the practice opportunities that would maintain the skill, while simultaneously creating a dependency that makes the skill essential during system failure. The result is a structural mismatch between the human's maintained capabilities and the system's failure modes.

The phenomenon was first identified by Lisanne Bainbridge in her analysis of the ironies of automation, where she observed that automated systems systematically degrade the human capabilities they nominally depend on for backup. A pilot who flies only on autopilot loses manual handling skills. A surgeon who uses only robotic assistance loses tactile diagnostic sensitivity. The system does not merely replace the human; it consumes the human capital required for its own safe failure.

Skill atrophy interacts catastrophically with automation complacency and out-of-the-loop unfamiliarity. The operator who has not practiced the manual skill for months is not merely rusty; they are operating in a different cognitive regime. Their mental models have decayed, their reflexes have slowed, and their confidence has become decoupled from their actual competence. When the system fails and demands manual intervention, the operator is structurally unprepared.

The resilience engineering response is structured manual practice — periodic handover of control to human operators not because the system has failed but because the operator's skills must be maintained. This is not a training problem; it is a design problem. Systems must be designed to require human engagement, not merely to tolerate it. The feedback topology of a safe system includes loops that keep the human operator competent, not merely present.