Human Factors

SECTION THREE

Human Performance Optimization

CHAPTER 16 | HUMAN FACTORS

Human Performance, Fatigue, Error, Anthropometry, CRM & Unmanned Aviation — CAsP Unit 16 Objectives 1–12

Human factors is the scientific discipline concerned with understanding interactions between humans and the elements of systems they operate, and applying that understanding to design and operations in order to optimize human well-being and overall system performance. In aerospace medicine, human factors is not a peripheral specialty — it is the integrating framework within which all physiological and psychological knowledge is applied to the built environment of the cockpit, the flight deck, and the operational setting. Aviation accidents are predominantly human-error events: estimates consistently attribute 70–80% of all aviation accidents at least in part to human error. Understanding why skilled professionals make errors, what conditions increase error probability, and how system design and crew training can reduce error consequences is the core mission of aviation human factors.

This chapter covers all 12 CAsP Unit 16 objectives: common human performance characteristics (16.1), fatigue in aviation (16.2), fatigue in space operations (16.3), human error in aerospace (16.4), historical error themes in accidents (16.5), cross-cutting factors contributing to error (16.6), anthropometry and operational requirements (16.7), cockpit mapping (16.8), user interface design (16.9), Crew Resource Management (16.10), unmanned vs. manned aviation human factors (16.11), and overall human factors objectives and components (16.12).

Human factors in aerospace is grounded in a set of fundamental human performance characteristics — the capabilities and limitations of the human cognitive, perceptual, and physical system — that determine how people interact with aerospace systems. These characteristics are not unique to aviation, but the aviation environment places them under unusual stress:

Attention and Working Memory

• Limited attentional capacity: Human attention is a finite, shareable resource. Divided attention (performing multiple concurrent tasks) degrades performance on all tasks as the number and complexity of simultaneous demands increases. Aviation tasks routinely require simultaneous monitoring of multiple systems, communication, navigation, and aircraft control.
• Working memory limitations: Working memory — the short-term, active cognitive workspace — holds approximately 7±2 chunks of information simultaneously and retains them for only seconds without rehearsal. In the cockpit, excessive information loading, interruptions, and distractions can cause information held in working memory to be lost before it is acted upon.
• Inattentional blindness: When attention is fully occupied by a task, stimuli that fall directly on the retina may not reach conscious processing (see Chapter 10, Objective 10.9). A pilot focused on avionics programming may fail to detect a threat entering the visual field.

Perception and Situation Assessment

• Expectation bias (confirmation bias): Humans tend to perceive what they expect to perceive. Once a mental model is established (e.g., ‘the runway ahead is clear’), contradicting information may be filtered, discounted, or misinterpreted to fit the expectation. This is a major source of errors in both cockpit and air traffic control operations.
• Pattern recognition and heuristics: Experienced operators recognize patterns rapidly and apply learned solutions (heuristics). This expertise is operationally efficient but can lead to errors when a novel situation superficially resembles a familiar pattern — the expert applies the learned response to the wrong situation.

Decision Making

• Naturalistic (Recognition-Primed) Decision Making: Expert decision makers rarely engage in analytical comparison of options. Instead, they recognize the situation from experience, generate a single plausible course of action, mentally simulate its execution, and implement it if it seems workable. This approach is fast and generally effective but is vulnerable to situations that violate the operator’s experience base.
• Decision under time pressure: As available decision time decreases (e.g., short time to impact, rapid aircraft upset), decision quality typically degrades and the tendency to revert to trained (automatic) responses increases. Time-critical emergencies favor automatic responses over analytical deliberation.

Automation and Complacency

• Automation bias: Operators tend to over-rely on automated systems, accepting automated outputs without independent verification. This can lead to failure to detect automated system errors until consequences are severe.
• Skill degradation (out-of-the-loop): Highly automated cockpits reduce the frequency with which pilots manually fly the aircraft. Manual flying skills (particularly during approach and unusual attitude recovery) can degrade without regular practice, reducing capability precisely when automation fails and manual skills are most needed.

Fatigue is a physiologically and cognitively degraded state arising from extended wakefulness, sleep deprivation, circadian disruption, or sustained physical or mental effort. In aviation, fatigue is not merely an inconvenience — it is a flight safety hazard directly comparable in effect to legal alcohol intoxication at moderate to severe levels.

Sources of Fatigue in Aviation

• Cumulative sleep debt: Commercial and military aviators routinely accumulate sleep deficits across duty periods and across weeks of rotating schedules. Even modest chronic sleep restriction (sleeping 6 hours vs. 8 hours per night) accumulates a performance debt equivalent to 1–2 nights of total sleep deprivation within 2 weeks.
• Circadian disruption (jet lag and shift work): Transmeridian travel (jet lag) and rotational shift schedules desynchronize the internal circadian clock from the local sleep-wake schedule. Performance is particularly degraded when alertness demands coincide with the circadian nadir of core temperature (approximately 0300–0600 local time).
• Extended duty periods: Long duty periods (>12–16 hours) accumulate wakefulness time to levels where cognitive performance deterioration is severe. 17–19 hours of continuous wakefulness produces cognitive deficits equivalent to a blood alcohol concentration of 0.05–0.08%.
• Low-stimulation conditions: Long-duration cruise flight, night operations with minimal activity, and monotonous cruise at high altitude produce ‘micro-sleep’ episodes (2–5 second involuntary sleep intrusions during apparent wakefulness) at fatigue levels that pilots may not subjectively recognize.

Aviation Regulatory Framework

Fatigue management in commercial aviation is regulated by crew rest and duty time rules that limit flight hours per day, week, and month and mandate minimum rest periods. The FAA, ICAO, and individual military services all maintain their own fatigue management standards. These rules are based on the cumulative fatigue data but are imperfect — individual variability in fatigue tolerance is large, and regulatory rest may be insufficient for some individuals under some conditions.

Space operations present a unique fatigue environment that combines nearly all of the fatigue-producing factors from aviation with several space-specific hazards:

• 16 sunrises per day on ISS: An astronaut aboard the International Space Station experiences a day-night cycle of approximately 92 minutes (16 complete cycles per 24-hour Earth day). Without external zeitgebers (time-givers such as natural light cycles) to entrain the circadian clock, astronauts must maintain a prescribed sleep-wake schedule anchored to Mission Control’s time zone. This artificial schedule fights against the natural tendency of the circadian clock to respond to the non-24-hour light cycles outside the spacecraft.
• Noise and environmental disruption: ISS environmental noise from life support systems, fans, pumps, and exercise equipment consistently exceeds NASA’s own guideline levels. Sleep quality is substantially degraded by noise, thermal discomfort, and microgravity-related difficulty finding comfortable sleep positions.
• Work overload and irregular scheduling: Mission timelines frequently run long, compressing scheduled sleep periods. Crew members on ISS missions routinely obtain less sleep than scheduled (typically 6.0–6.5 hours rather than the recommended 8.0 hours), accumulating chronic sleep restriction.
• Circadian disruption during launch, transit, and landing: The extreme schedule changes of launch preparation, the transition to microgravity, and the re-entry and landing sequence all produce acute circadian disruption at safety-critical mission phases.
• Pharmacological sleep aids in space: A significant proportion of long-duration mission crew members use prescribed sleep aids (typically zolpidem or other hypnotics) to manage sleep difficulties in the space environment. These medications carry risks of residual sedation, impaired memory consolidation, and ‘sleep inertia’ (post-awakening grogginess) — potentially impaired performance during emergency night-time operations.
• Countermeasures: Scheduled light exposure (timed to anchor circadian phase), scheduled naps, physical exercise (which promotes deeper sleep architecture), and behavioral sleep hygiene measures. Research into optimized light wavelength and timing for circadian anchoring on ISS is ongoing.

Human error in aerospace is a systemic phenomenon — not a personal failing. When skilled, experienced, and motivated professionals make errors while performing tasks that are not considered terribly difficult, the cause is rarely simple negligence or incompetence. Rather, errors arise from the intersection of: inherent limitations of human cognitive processing, task demands that exceed cognitive capacity, environmental factors (fatigue, stress, distraction, time pressure), and organizational factors (procedures, training, culture, design) that create ‘error-prone’ conditions.

Prevalence and Attribution

Estimates consistently attribute 70–80% of all aviation accidents at least in part to human error. However, attributing an accident to ‘pilot error’ without analyzing the underlying cognitive and organizational factors that created the error opportunity is operationally useless — it identifies what failed without explaining why or how to prevent recurrence.

James Reason’s Error Typology

James Reason’s influential taxonomy categorizes errors by the level of cognitive processing at which they occur:

• Slips: Errors of action arising from failures of execution (not intent). The correct intention was formed, but the action sequence diverged from intention during automatic, skill-based execution. Example: Setting the wrong altitude on the autopilot while correctly intending to set the correct altitude — a miskey during a habitual motor sequence.
• Lapses: Failures of memory. The correct intention was formed, but the required action was forgotten before completion. Example: Forgetting to complete a checklist item after an interruption.
• Mistakes: Errors of planning or decision-making. The intended action was executed correctly, but the plan or decision itself was wrong. These are knowledge-based or rule-based failures: the wrong rule was applied (rule-based mistake) or insufficient knowledge led to a flawed analysis (knowledge-based mistake).
• Violations: Deliberate deviations from rules, procedures, or standards. Unlike errors (which are unintentional), violations are intentional acts. They may be routine (normalized deviance, where violations become accepted practice), situational (context-specific decision to deviate), or exceptional (rare, deliberate breach). Violations are not necessarily malicious but reflect organizational failures in procedure design, culture, and enforcement.

The Swiss Cheese Model

Reason’s Swiss Cheese Model of accident causation describes accidents as the result of failures at multiple layers of defense simultaneously aligning: each layer of organizational defense (procedures, training, equipment design, supervision) has holes (latent failures or active failures). An accident occurs when holes in multiple layers align, creating a trajectory from hazard to harm. Active failures are the immediate human errors; latent failures are the pre-existing organizational conditions that create the vulnerability for active failures to produce accidents.

Analysis of aviation accident databases has identified recurring thematic categories of human error that appear across accident types, aircraft categories, and time periods. These themes represent the highest-priority targets for human factors intervention:

Table 16.1. Historical Human Error Themes in Aviation Accidents

Cross-cutting factors are conditions or states that increase human error vulnerability across all types of tasks and all phases of flight — they are not specific to one error type but increase the probability of any error. These are sometimes organized using the SHELL model or the PEAR model in aviation human factors:

Cross-Cutting Error Vulnerability Factors

• Fatigue and sleep deprivation: The most consistently documented performance degrader in aviation. Reduces sustained attention, working memory capacity, reaction time, risk perception accuracy, and decision quality. As described in Objective 16.2, 17–19 hours of wakefulness produces deficits equivalent to 0.05–0.08% BAC. Critically, fatigued operators consistently underestimate their own impairment.
• Stress (acute and chronic): Acute stress narrows attention (attentional tunneling — beneficial for processing the most salient threat but harmful for monitoring lower-salience but important information). Chronic stress from workload, life events, or organizational factors degrades baseline cognitive capacity and motivation.
• Time pressure: Reduces deliberation time available for decision-making, increasing reliance on heuristics and first-available responses. High time pressure degrades decision quality, particularly for novel or complex situations requiring systematic analysis.
• High workload: Exceeds available cognitive capacity, producing task-shedding (the conscious or unconscious dropping of lower-priority tasks) and increased error probability for all concurrent tasks.
• Distraction and interruption: Interruptions during procedural tasks (checklists, programming) are a major source of lapse errors — the operator returns to the task but loses their place, omitting steps. Distraction from external sources (passengers, ATC, maintenance alerts) can disrupt attention during critical task phases.
• Complacency (automation-induced): As described in Objective 16.1, over-reliance on automation reduces monitoring vigilance and allows automated errors to propagate before detection.
• Poor communication: Ambiguous ATC instructions, inadequate crew briefings, unclear authority gradients, and failure to assert safety concerns are common factors in crew coordination accidents.
• Inadequate training or currency: Skills not recently practiced are less reliable under pressure. Rare, abnormal situations — for which training was less thorough — are disproportionately represented in accidents.
• Environmental stressors: High noise, thermal stress, hypoxia, vibration, and confined workspace all degrade cognitive performance and contribute to error vulnerability as described in previous chapters.

Anthropometry is the scientific measurement of the human body’s physical dimensions and characteristics, and its application to the design of equipment, workspaces, and operational systems. In aerospace, anthropometry ensures that aircraft, spacecraft, and support equipment can be safely and effectively operated by the intended user population.

Why Operational Requirements Drive Anthropometry

An operational requirement specifies what tasks the human must be able to perform in the system, under what conditions, and to what standard of performance. Anthropometric design is derived FROM the operational requirement — not the other way around:

• Defining the user population: The first step is defining the range of human body dimensions that must be accommodated. Military aviation typically designs to accommodate the 5th through 95th percentile of the target population in critical dimensions. Designing only for the mean (50th percentile) would exclude approximately 50% of potential operators.
• Functional fit requirements: The operational requirement defines the specific functions that must be accommodated: Can the pilot reach all controls? Can they see all displays? Can they perform the ejection seat handle pull? Can they don and doff the equipment within required time limits? Can they survive a 14-G crash impact? These functional requirements — not just static measurements — are the criteria for anthropometric design.
• Dynamic considerations: Static (seated) anthropometry is insufficient. Operational requirements include dynamic scenarios: reach under high G, field of view through NVG helmet, manual dexterity in pressurized gloves, and egress time from a damaged aircraft. A pilot who meets static fit criteria may not meet dynamic functional criteria — particularly relevant for women entering tactical fighter aviation.
• Ejection seat compatibility: One of the most critical anthropometric operational requirements in aviation. Safe ejection requires body mass within design limits (too light → inadequate ballistic support; too heavy → excessive spinal loading), seated height within a defined range, and appropriate positioning in the seat. Body mass and height limits for ejection seats are explicit anthropometric operational requirements that determine pilot flight eligibility.

Cockpit mapping is the process of systematically determining whether the body dimensions of a target user population allow them to safely reach, see, operate, and egress the cockpit and its controls and displays — and documenting the range of users who CAN and CANNOT be accommodated.

Process of Cockpit Mapping

Dynamic considerations in cockpit mapping: Static reach in a seated position is not sufficient. Research has documented that pilot reach to certain controls is substantially reduced under high positive G (cockpit reach under G was directly studied using a centrifuge). The dynamic environment — high G, pressure suit restriction, NVG weight and center of mass, glove restriction — must be included in the mapping assessment.

A dramatic practical example: A 12-lb head in a 1-G environment weighs more than 100 lb in a 9-G turn. A 4-lb NVG/helmet system in 1-G becomes 36 lb in a 9-G turn. The static accommodation of a helmet that ‘fits’ in 1 G may produce dangerous movement and reduced field of view under 9-G tactical maneuvering.

A user interface is anything a user interacts with — computer, display, control, tool, procedure card, or any other system element through which the human passes information to or receives information from the system. User interface design is one of the most impactful areas of applied human factors because a well-designed interface reduces error probability, reduces training requirements, and reduces operator workload, while a poorly-designed interface creates the preconditions for errors.

Usability: The Central Design Goal

The key concept in user interface design is usability — the degree to which a system can be used by specified users to achieve specified goals with effectiveness, efficiency, and satisfaction. A usable interface is:

• Useful (does what the user needs it to do).
• Usable (can be used by the intended operator, with minimal training and errors).
• Easy to learn and remember.
• Safe (produces fewer errors and makes errors recoverable before consequences develop).
• Subjectively satisfying (comfortable to use, reducing stress and fatigue).

Principles of Good User Interface Design

• Early and continuous focus on users: Design must be driven by the actual needs, capabilities, and limitations of the intended users — not by engineering convenience or aesthetic preference. User requirements must be established before design begins.
• Consistency: Similar controls should operate similarly; similar displays should look similar. Inconsistent interfaces increase cognitive load and error probability.
• Visibility: The current state of the system should be visible to the operator. Hidden system states (e.g., autopilot mode annunciation that is small or easy to miss) create mode confusion errors.
• Feedback: Every user action should produce clear, immediate feedback confirming that the action was recognized and executed. Absent or delayed feedback promotes repeated incorrect inputs.
• Error prevention and recovery: The interface should make errors difficult to commit (interlocks, confirmation requirements) and should make errors recoverable before consequences develop.
• Iterative testing: User interface design requires repeated test-modify-retest cycles with representative users. No interface design is correct without empirical validation with real users in representative conditions.

Crew Resource Management (CRM) is formal training for aircrew designed to improve crew coordination, communication, and decision-making, with the ultimate goal of improving flight safety through effective management of all available resources: human resources, hardware, and information.

Historical Origin

CRM originated from a 1979 NASA workshop on flight crew error that identified crew coordination failures (rather than lack of technical skill) as a dominant factor in commercial airline accidents. Analysis showed that up to 70% of accidents involved crew issues rather than mechanical, maintenance, or weather factors. The term ‘cockpit resource management’ (later expanded to ‘crew’ to include cabin crew and other personnel) was born.

Core CRM Competencies

• Communication: Clear, unambiguous transmission and reception of information. Closed-loop communication (sender → receiver → read-back → confirmation) to verify accuracy. Assertive communication — the obligation of any crew member to speak up whenever safety is threatened, regardless of rank or authority.
• Situational awareness: Accurate perception of current flight conditions (aircraft state, weather, traffic, systems status, terrain). Shared situational awareness — the entire crew shares a common accurate mental model.
• Decision making: Structured approaches to problem identification, option generation, and course-of-action selection under time pressure. Recognition-Primed Decision Making for time-critical scenarios; analytical approaches for complex novel situations.
• Leadership and authority management: Clear establishment of decision-making authority appropriate to the situation. The Pilot in Command (PIC) bears ultimate responsibility but must create an environment where other crew members can freely raise safety concerns. Excessive authority gradient (the PIC is always right; subordinates don’t speak up) is a documented accident factor.
• Task management and workload management: Prioritizing tasks according to urgency and importance; delegating appropriately to reduce individual cognitive overload; managing the transition between normal, abnormal, and emergency workload states.
• Threat and Error Management (TEM): The current state-of-the-art in CRM. Accepts that threats and errors are inevitable; focuses on proactive identification of threats (conditions that increase error probability), avoidance of errors where possible, trapping errors before they produce adverse outcomes, and managing error consequences when trapping fails.

Effectiveness of CRM Training

CRM training is well-established in commercial aviation (required by FAA Advisory Circular AC 120-51E for all U.S. commercial carriers) and in military aviation. Research supports that well-designed CRM programs improve crew coordination behaviors measured in simulation. The evidence for reduction in accident rates specifically attributable to CRM is more difficult to establish due to the multifactorial nature of accident causation, but CRM is now considered a core component of aviation safety infrastructure.

Unmanned aircraft systems (UAS) represent a rapidly growing proportion of both military and civil aviation operations. While they eliminate the risk of aircrew dying in the aircraft, they introduce a distinctive set of human factors concerns that differ substantially from those of manned aviation:

Table 16.2. Human Factors Comparison: Manned vs. Unmanned Aviation

Key Objectives of Human Factors

As stated by Chapanis, human factors engineering is the application of human factors to the design of systems, machines, tools, tasks, and environments for safe, effective, and comfortable human use. The ultimate objectives are:

• Facilitating operational efficiency: Increase safety; minimize error; increase system reliability.
• Achieving reliability, maintainability, and availability: Human-centered design produces systems that are easier to maintain correctly, fail more gracefully, and recover faster from failures.
• Providing user-centered design: Improve the work environment; reduce stress and fatigue; increase ease of use, comfort, and user acceptance. Place the human’s capabilities and limitations at the center of the design process.
• Reducing collateral costs: Reduce loss of time, equipment, and human life attributable to design-induced errors and human-system mismatches.

Key Components of Human Factors

Table 16.3. Key Components of Human Factors in Aerospace Medicine

A multidisciplinary team is required to apply human factors with a systems approach. This team may include experts in: psychology, industrial engineering, occupational medicine, applied physiology, anthropometry, industrial design, operations research, and statistics. Aviation physiology contributes directly to such teams through expertise in physiological performance limitations (hypoxia, G-forces, spatial disorientation, fatigue, thermoregulation, dehydration) that are directly relevant to human factors analysis.