HFACS (Human Factors Analysis and Classification System)

The Human Factors Analysis and Classification System (HFACS) was developed by Dr. Scott Shappell and Dr. Douglas Wiegmann around the turn of the 2000s — Shappell then at the U.S. Civil Aeromedical Institute and Wiegmann at the University of Illinois — and set out most fully in their 2003 book A Human Error Approach to Aviation Accident Analysis. It was created to solve a practical problem: investigations repeatedly concluded that human error was the primary causal factor in the large majority of accidents, but "human error" as a finding is analytically useless because it does not say what kind of error, why it occurred, or what could be changed to prevent the next one. HFACS gives investigators a common, reliable taxonomy for answering those questions.

HFACS is the operational counterpart to James Reason's Swiss cheese model. Reason described accidents as the alignment of failures across successive defensive layers but did not provide a codeable scheme for classifying what those failures actually are. HFACS fills that gap by defining four levels of failure, each with specific sub-categories, mapped directly onto Reason's layers. Level 1, unsafe acts, covers the actions of the front-line operator and divides into errors (further split into decision errors, skill-based errors, and perceptual errors) and violations (routine and exceptional). Level 2, preconditions for unsafe acts, addresses the conditions that set up those acts — including the condition of the operator (adverse mental states, adverse physiological states, and physical or mental limitations), personnel factors (crew resource management and personal readiness), and environmental factors (the physical and technological environment). Level 3, unsafe supervision, examines the supervisory chain: inadequate supervision, planned inappropriate operations, failure to correct a known problem, and supervisory violations. Level 4, organizational influences, reaches the highest and most latent layer: resource management, organizational climate, and organizational process.

The defining value of HFACS is that it forces the analysis upward, past the operator. Because the framework requires investigators to consider all four levels, it makes latent organizational contributors visible instead of allowing an investigation to close at the sharp end with a finding of pilot or technician error. A skill-based error at Level 1 might be linked to fatigue at Level 2, which is linked to a planned inappropriate roster at Level 3, which is linked to a resource-management shortfall at Level 4 — and only by naming all four does the investigation identify a fix that will actually reduce recurrence. This is the same insight that underpins just culture: the person who committed the unsafe act is usually the inheritor of latent conditions created higher up, not the sole cause.

Beyond single-event investigation, HFACS is a powerful tool for trend analysis. Because it produces consistent, categorized data, an organization or authority can aggregate HFACS codes across many events and identify which sub-categories recur — revealing, for example, that supervisory violations or a particular precondition are a systemic pattern rather than one-off occurrences. This makes it valuable to safety management systems as a way of turning a body of occurrence reports into targeted, prioritized safety action. Variants have been developed for specific domains, including HFACS-MX for aircraft maintenance, and the framework is widely used by military and civil investigators, regulators, and researchers. HFACS is complementary to, rather than a replacement for, operator-side investigation tools such as MEDA in maintenance: MEDA captures the contributing factors of a specific error immediately after it happens, while HFACS provides the classification structure for analysis and trending across many events. It also sits naturally alongside flight-data-derived programs such as FOQA/FDM and observation programs such as LOSA, giving a human-factors classification to the events those programs surface.

For a training organization or combined ATO/AOC operator building the safety-assurance side of its SMS, HFACS provides the missing link between collecting occurrence reports and actually learning from them. A pile of reports each closed with "instructor error" or "student error" teaches an organization very little; the same reports classified through the four HFACS levels reveal whether the real drivers are, for instance, adverse mental states from scheduling pressure (a Level 2 precondition) or a failure to correct a known problem (a Level 3 supervisory factor). That distinction is the difference between telling people to be more careful and fixing the roster or the supervisory process.

HFACS also reinforces just culture in a concrete, procedural way. By building the search for supervisory and organizational contributors into the investigation method itself, it prevents the analysis from defaulting to blame at the front line and directs corrective action to the latent conditions the organization actually controls. For maintenance-heavy operators, the HFACS-MX variant gives the maintenance department the same structured lens, so that a mis-installation is analyzed for its supervisory and organizational contributors rather than being attributed to a single technician's mistake.

Aviatize's Safety Management module lets an organization classify each occurrence and investigation against a structured taxonomy of contributing factors aligned with the HFACS levels, storing the classification as data rather than free text. That makes aggregation possible: the platform can surface that a particular precondition or supervisory factor is recurring across many otherwise-unrelated events, which is the signal that a systemic corrective action — not another individual reminder — is required. Findings link to risk assessments and corrective-action tracking so the loop from analysis to fix is closed and auditable.

Because the safety data connects to the KPI Reporting & Dashboards module, the accountable manager sees HFACS-based trends over time as safety-performance indicators, and the Compliance & Auditing module keeps the investigation records and their organizational-level findings available as evidence of a functioning safety-assurance process.