What is a Line Operations Safety Audit (LOSA) and how is it different from a line check?

A LOSA is a voluntary, non-punitive observation of normal revenue flights by a trained independent observer who records how crews manage threats, errors, and undesired aircraft states. Unlike a line check, LOSA data is collected anonymously, cannot be used for enforcement or discipline, and produces aggregate organizational trends rather than individual evaluations. Its sole purpose is organizational learning and SOP improvement, as outlined in AC 120-90.

What does normalization of deviance mean in aviation?

Normalization of deviance is the gradual process by which repeated, uncorrected departures from standard operating procedures come to be accepted as normal by a crew or organization. Each small deviation raises the threshold for noticing the next one, so drift from the original standard accumulates without triggering alarm. Aviation accident investigations have repeatedly found that crews involved in accidents had long-established patterns of accepting non-standard practices that eroded the safety margins SOPs were designed to preserve.

How does LOSA help prevent normalization of deviance in airline operations?

LOSA helps detect normalization of deviance by placing an independent, trained observer on normal flights who can objectively record deviations that crews themselves no longer perceive as abnormal. When aggregated LOSA data reveals that a high percentage of crews are consistently deviating from a particular SOP, the airline can investigate whether the procedure is unclear, whether training is reinforcing the wrong behavior, or whether crew culture has silently endorsed the shortcut — and then take corrective action before the pattern contributes to an accident.

Line Operations Safety Audit (LOSA) and Normalization of Deviance

Aviation safety has long relied on accident and incident investigation to find what went wrong after the fact. But two complementary concepts — the Line Operations Safety Audit (LOSA) and the phenomenon known as normalization of deviance — shift the lens forward, seeking to identify hazards and eroding standards before an accident occurs. For Airline Transport Pilot candidates and professional flight crews, understanding both concepts is essential not only for knowledge-test success but for building a safety culture that genuinely prevents catastrophic outcomes.

LOSA is a structured, voluntary, and non-punitive observation program in which trained observers ride along on normal revenue flights and systematically record how crews manage threats, errors, and undesired aircraft states. Normalization of deviance, by contrast, is a sociological concept adopted by aviation safety professionals to describe the gradual process by which deviation from standard operating procedures (SOPs) becomes accepted as normal within a group or organization. Together, they form a powerful framework for proactive safety management under the FAA's Safety Management System (SMS) philosophy.

How LOSA Works

According to Advisory Circular 120-90, LOSA is built on direct, jump-seat observation of line operations by trained observers who are independent of management authority and who collect data anonymously. The program is specifically designed so that no individual crew member can be identified in the final data set — this confidentiality is the cornerstone that motivates crews to behave naturally rather than performing for an evaluator. The data collected is not used for enforcement or disciplinary action; its sole purpose is organizational learning.

Observers use structured data-collection worksheets to capture three broad categories of crew performance drawn from the Threat and Error Management (TEM) model:

Threats: Environmental or operational hazards that exist before the crew acts — weather, ATC complexity, maintenance issues, or challenging airports.
Errors: Actions or inactions by the crew that deviate from organizational intentions or SOP requirements, whether or not they result in consequences.
Undesired Aircraft States (UAS): Aircraft positions, speeds, configurations, or attitudes that result from errors or threats and that, if unresolved, increase risk — for example, unstabilized approaches or incorrect flap configurations.

After a sufficient sample of observations is gathered — AC 120-90 recommends the sample be statistically robust and representative of the carrier's route structure and fleet — the data is aggregated and analyzed. The output is not a list of rule violators but a profile of systemic vulnerabilities: which threats recur most often, which types of errors are most prevalent, and which SOPs show the highest rate of deviation. This systemic view is what distinguishes LOSA from a traditional line check.

The LOSA Observer and the Role of Confidentiality

Observers are typically experienced pilots trained specifically in LOSA methodology. They introduce themselves to the crew, explain the voluntary and non-punitive nature of the program, and obtain consent. If a crew declines to participate, no data is recorded and no adverse notation is made. This voluntary aspect reinforces the trust that is indispensable to the program's integrity. The LOSA methodology was developed by human factors researchers at the University of Texas and later documented and formalized by the International Civil Aviation Organization (ICAO) in Document 9803, and the FAA codified guidance for U.S. air carriers in AC 120-90.

Normalization of Deviance

Sociologist Diane Vaughan introduced the term normalization of deviance in her book analyzing the Space Shuttle Challenger disaster, and aviation safety experts quickly recognized its direct applicability to flight operations. Normalization of deviance occurs when a small departure from a standard — say, consistently beginning a descent checklist slightly after the initial descent call rather than at it — goes uncorrected and unnoticed long enough that the crew, the team, or the organization begins to perceive it as acceptable. Over time, the deviant behavior becomes the de facto standard.

The danger is not any single deviation; most individual shortcuts carry small risk. The danger is the trajectory: as one deviation is accepted, the threshold for noticing the next one rises, and the process repeats. Each step away from the SOP feels small relative to where the crew currently stands, even though the cumulative drift from the original standard may be substantial. Aviation accident research has found this pattern repeatedly — crews who had routinely flown unstabilized approaches to visual landings, skipped callouts during abnormal checklists, or briefed approaches at altitudes below those prescribed in SOPs, eventually encountered conditions where that accumulated drift had eliminated the safety margins the procedures were designed to preserve.

LOSA as a Tool to Detect Normalization of Deviance

LOSA's greatest contribution to combating normalization of deviance is its ability to make the invisible visible. Because individual crews often do not recognize their own drift from standards — the deviation feels normal to them — an independent, structured observer can capture what internal check rides or self-reporting systems miss. When LOSA data shows that 40% of observed crews are consistently skipping a particular callout, or that a specific transition altitude is regularly being used incorrectly on a given route, the organization has actionable, system-level information. Leadership can then investigate whether the SOP is ambiguous, whether training reinforced the correct behavior, or whether peer culture has silently endorsed the shortcut.

Why It Matters Operationally

For line crews, the importance of LOSA and normalization of deviance concepts extends beyond passing an ATP knowledge test. Part 121 air carriers operating under FAA-accepted SMS frameworks are increasingly expected to conduct proactive safety assessments of normal operations — LOSA is one of the most validated tools available. For the individual pilot, awareness of normalization of deviance is a professional defense mechanism: recognizing that comfort with a practice is not the same as safety of the practice is a critical metacognitive skill.

Consider the stabilized approach: individual operators' FAA-accepted Operations Specifications and SOPs — commonly built around FAA guidance such as AC 120-71 — establish stabilized approach criteria, with many carriers requiring stabilization by 1,000 feet AGL in IMC and 500 feet AGL in VMC, along with specific criteria for airspeed, sink rate, configuration, and power. If a crew consistently lands successfully from approaches that were not technically stabilized at those gates, success gradually normalizes the deviation. LOSA data from multiple carriers has historically identified high rates of approach-and-landing deviations as the single most common undesired aircraft state — directly reflecting the cumulative effect of normalization of deviance across fleets.

Key Numbers and Rules

AC 120-90: FAA Advisory Circular establishing LOSA guidance for Part 121 air carriers; based on the LOSA methodology documented in ICAO Document 9803.
Confidentiality requirement: LOSA data must be collected anonymously; no individual identification in reports; not used for enforcement or discipline.
Voluntary participation: Crews must consent; refusal carries no penalty and no record.
TEM model categories: Threats, Errors, and Undesired Aircraft States — the three-part taxonomy used on LOSA observation worksheets.
Stabilized approach gates (common industry practice): 1,000 ft AGL in IMC, 500 ft AGL in VMC, as commonly established by individual carriers' FAA-accepted OpSpecs/SOPs — frequently identified in LOSA data as a common UAS boundary.
LOSA purpose: Organizational learning and SOP improvement, not individual evaluation.
SMS connection: LOSA data primarily supports the Safety Assurance component of the four-pillar SMS model, and findings can also inform Safety Risk Management (Safety Policy, Safety Risk Management, Safety Assurance, Safety Promotion).

Common Test Traps

LOSA vs. line check confusion: A LOSA observation is not an evaluation and cannot be used for enforcement — test questions may imply it functions like an IOE or proficiency check. It does not.
Normalization of deviance is gradual, not sudden: Examiners may frame it as a single decision or conscious choice. In reality it is an incremental, often unconscious organizational drift.
LOSA data is aggregate, not individual: The program produces fleet-wide trends, not a report card on specific pilots. Candidates sometimes confuse LOSA with ASAP (Aviation Safety Action Program), which does involve individual voluntary reports.
Threats are not errors: In the TEM model, a threat (e.g., a bird strike, an ATC frequency change at a busy moment) exists independently of crew action. An error is a crew action or inaction. Conflating them is a common exam mistake.
Successful outcomes do not validate deviant practices: A crew that consistently lands safely from an unstabilized approach has not proven the approach was safe — they have only demonstrated they got away with it, which is the core mechanism of normalization of deviance.

Taken together, LOSA and the concept of normalization of deviance represent the proactive face of modern aviation safety management. LOSA gives organizations the data to see systemic drift that crews cannot see from inside it, and understanding normalization of deviance equips individual pilots to question their own comfort with non-standard practices before that comfort creates a vulnerability. For the ATP candidate, both concepts reflect the broader human factors principle that safety is not simply the absence of accidents — it is the active, continuous management of risk in the gap between the way things are supposed to be done and the way they are actually done.