Cognitive limits, bias and training: what really needs to change in the simulator

By Captain Bassani - ATPL/B-727/DC-10/B-767 - Former Air Accident Inspector SIA PT. https://www.personalflyer.com.br - captbassani@gmail.com - Mar/2026

Image AI

Training pilots today means training the brain that runs the cockpit: cognitive workload, bias, and automation must appear explicitly in simulator briefs, exercises, and debriefs, not only between the lines of CRM slides. Studies using HRV, EEG and performance measures in commercial jet simulators show that takeoff and landing concentrate peak mental workload, higher error propensity and greater vulnerability to decision biases, while the cruise phase tends to favor physiological and cognitive recovery.

From “flying the profile” to “managing the brain”

In aircraft such as the A320 and B737, experimental work shows that as tasks accumulate on approach, vectors, runway changes, deteriorating weather, system messages, automation mode changes, pilots gradually drop peripheral tasks to preserve basic flight path control. This leads to more reading errors, reduced cross‑monitoring, and delayed critical decisions such as the go‑around.​

At the same time, occurrence reports and accident analyses repeatedly identify plan continuation bias (get‑there‑itis), confirmation bias, and outcome bias in decisions to continue unstable approaches, press on into marginal weather, or delay go‑arounds despite multiple cues that the plan is no longer safe. In this context, the goal of simulator training cannot be limited to “flying the published profile”; it must also develop the ability to recognize when cognitive limits have been reached, identify the biases that are distorting situational awareness, and choose concrete operational actions (simplify the plan, execute a go‑around, request vectors, divert).

What operators should demand from simulator training and scenarios that push workload, not only rehearse the normal

Modern LOFT and EBT programs should explicitly include:

• Approach and takeoff profiles designed to generate high workload, combining late vectors, runway changes, deteriorating weather, minor technical issues and ATC constraints, until simplifying the plan (go‑around, holding, diversion) clearly becomes the safest and most professional option.​

• Objective workload and performance measures, such as response times to alerts, frequency of reading errors, trajectory instability, number of PM intervention calls, and, where feasible in research or advanced programs, tools like NASA‑TLX, HRV, or secondary‑task reaction times.​

Cognitive bias as a deliberate ingredient

Instead of treating biases as abstract classroom concepts, the simulator should deliberately trigger them so that pilots can recognize them in practice:

• Plan continuation bias: scenarios where a combination of instability, wind, weather, and profile deviations makes a go‑around the correct choice – and debriefs where competence is explicitly credited to the early decision to discontinue, not to “making it fit.”​

• Confirmation bias: situations that start with optimistic information (briefing, TAF, expected runway) and progressively introduce meteorological discrepancies, runway changes or system limitations, forcing the crew to acknowledge that the initial mental model no longer holds and to re‑brief and re‑plan.​

• Debriefs centered on the decision-making process: analyzing what the crew knew at the time, how threats were prioritized, and which alternatives were considered, rather than judging the session solely by its outcome. This approach reduces outcome bias and supports learning from near-miss events.

Workload Management as an observable competency

EBT and CBT frameworks already list Workload Management, Situational Awareness, and Automation Management as core competencies, but they often remain at a generic level. To make them operationally meaningful, operators should:

• Define clear behavioural markers, such as anticipating FMS programming, redistributing PF/PM tasks at workload peaks, deferring non‑essential activities in critical phases, reducing or simplifying automation when it starts consuming more attention than it saves, and calling a tactical time‑out when the situation becomes overwhelming.

• Document in checks and EBT sessions whether the pilot protected critical phases from unnecessary distractions (late briefings, complex reprogramming, side discussions), requested help in time, and chose the prudent option of discontinuing a compromised approach, even if a landing would still have been technically possible.

Universities, authorities and manufacturers: converging agendas

Research in neuroergonomics and simulator‑based training shows that it is feasible to monitor and classify pilots’ workload and cognitive control while flying traffic patterns, identifying moments of saturation and loss of cognitive control. International authorities and technical groups (ICAO, IATA, EASA, FAA) have been incorporating this perspective into human‑factors, EBT, and CRM guidance, recommending that:​

• Concepts of cognitive limits and bias be included in instructor and examiner training, so they can recognise signs of saturation (attentional narrowing, loss of cross‑monitoring, sudden increase in simple errors) and intervene not only technically but also cognitively.

• Cognitive biases be integrated into TEM (Threat and Error Management), treating plan continuation bias, confirmation bias, and outcome bias as internal threats requiring explicit mitigation strategies (tactical pause, check‑and‑challenge, plan reset, active PM involvement in questioning risky decisions).​

Manufacturers have increasingly embedded human factors principles, operational resilience, and workload management into their safety philosophies and training programmes. This is reflected in cockpit designs and training concepts aimed at reducing unnecessary complexity, improving automation mode transparency, and facilitating the recovery of situation awareness after unexpected events. However, without a structured training programme that deliberately develops cognitive skills and decision-making under realistic operational stressors, even a state-of-the-art simulator remains merely advanced hardware, operating well below its full potential as a proactive safety tool.

What this implies in practice for an airline

To turn cognitive limits and bias from theory into behavioural change, an operator’s training programme should ensure that its simulators:

• Include dedicated exercises that push workload and deliberately trigger plan continuation and confirmation bias, followed by deep decision‑focused debriefing.

• Assess Workload Management, Automation Management and Situational Awareness transparently, aligned with EBT/CBT, and not only profile adherence.

• Use objective data whenever possible, such as performance metrics and structured workload assessments (and, in research settings, physiological tools), to support feedback to pilots.​

Reinforce a culture in which calling a go-around, requesting vectors, or initiating a deliberate tactical pause (time-out) is recognized as expected professional behavior rather than as a sign of weakness. Ultimately, training cognitive limits and biases in the simulator means investing in the only truly universal redundancy in aviation: the human who decides when to simplify the plan, discontinue the approach, and keep safety above any operational pressure.

Ultimately, training cognitive limits and biases in the simulator is an investment in aviation’s only truly universal “redundancy”: the human being who decides when to simplify the plan, execute a go-around, and prioritize safety over any operational pressure.

Want to increase flight safety? Send this post to your peers!

Safe Flights!

Captain Luiz Bassani

Sources

International Regulatory and Advisory Publications

• ICAO Human Factors Training Manual (Doc 9683).

• ICAO Manual on Evidence-based Training (Doc 9995).

• ICAO Safety Management Manual (Doc 9859).

• EASA Part-FCL and Acceptable Means of Compliance on CBT/EBT.

• FAA Advisory Circulars on CRM, human factors, and automation.

Human Factors and Workload Research

• Publications in Aviation, Space, and Environmental Medicine; The International Journal of Aviation Psychology; Human Factors.

• Conference proceedings (e.g., HFES, ICAS, AIAA Human Factors symposia).

Accident Investigation Reports

• Analyses from NTSB, AAIB, ATSB, TSB (Canada), CENIPA focusing on approach biases, CRM failures, and decision errors.

WINGS OF KNOWLEDGE SERIES

Take your aviation knowledge to the next level, secure your copies today! The first and second volumes of this 18-book series are now available online in Portuguese and English. Don’t miss the chance to access the reference guides that pilots, instructors, and aviation professionals are already using to advance their skills and understanding.

They are now available for online purchase, in Portuguese and English, the first volumes of this 18-book series, which is already emerging as a reference for aviation professionals and enthusiasts:

Volume 1 – Introduction to Aviation and the Role of the Pilot

Volume 2 – Human and Physiological Factors in Flight Safety

Volume 3 – The Importance of Aeronautical Knowledge – Mai 2026

More than books, these guides represent a true transformation in the education and development of aviation professionals, from cadets to experienced pilots, instructors, safety managers, and dedicated enthusiasts.

They provide a structured path toward mastering the art and science of flight, integrating technical fundamentals, human factors, internationally recognized best practices, and a forward-looking perspective.