Introduction
Incorrect take-off performance data remains a persistent safety risk in modern aviation, despite the widespread adoption of sophisticated digital tools such as Electronic Flight Bags (EFBs) and advanced Flight Management Systems (FMS). While these technologies have significantly improved computational accuracy and operational efficiency, recent incidents show that fundamental vulnerabilities—particularly data-entry errors, inadequate cross-verification, and overreliance on automation—continue to erode safety margins during one of the most critical phases of flight.
A series of events between 2023 and 2025, including a high-profile tail strike involving a LATAM Boeing 777-300ER, illustrate how relatively simple errors can cascade through multiple layers of defence when not detected in time. These events reinforce a key safety insight: the risk is not eliminated by automation but transformed, requiring renewed emphasis on human performance, procedural discipline, and system design.
The Critical Nature of Take-Off Performance
The take-off phase is a uniquely demanding operational environment in which aircraft transition rapidly from ground roll to airborne flight. It is characterised by high workload, rapidly changing aerodynamic conditions, and limited opportunity for corrective action once the aircraft passes decision speed (V1). Accurate take-off performance data is therefore essential to ensure that the aircraft can accelerate within the available runway, achieve appropriate rotation speeds, clear obstacles safely, and maintain adequate climb performance in the event of an engine failure.
Even small discrepancies in performance inputs—such as aircraft weight, runway length, environmental conditions, or configuration—can disproportionately affect safety outcomes. The margins available during take-off are inherently narrow, particularly for long-haul, high-weight operations in demanding environmental or runway conditions.
Although digital tools have largely eliminated traditional calculation errors associated with manual performance charts, they remain fundamentally dependent on the accuracy of input data. As a result, operational risk has shifted to a “garbage-in, garbage-out” paradigm, where incorrect inputs can produce internally consistent yet unsafe outputs.
Automation: Shifting Rather Than Eliminating Risk
The introduction of EFBs and FMS-based performance tools has undoubtedly enhanced operational efficiency. However, these systems are highly sensitive to incorrect inputs, including aircraft weight, flap configuration, runway selection, and environmental parameters. When erroneous data is entered, the resulting calculations—though mathematically correct—may be operationally invalid.
A growing concern in the industry is the erosion of rigorous manual cross-checking practices. As automated systems consistently produce reliable outputs under normal conditions, there is a tendency for flight crews to accept these outputs with less scrutiny. This over-reliance can be compounded when both pilots independently enter the same incorrect data, or when values are communicated verbally and replicated, effectively contaminating what is intended to be an independent verification process.
Human factors play a central role in this dynamic. Expectation bias may lead crews to accept performance figures that “look about right” for a given operation, while confirmation bias reinforces acceptance of outputs that align with preconceived expectations. These tendencies are especially pronounced under time pressure, such as during rapid turnarounds at congested airports, where operational demands can compress decision-making timelines.
The growing complexity of airport environments further exacerbates these risks. Frequent runway changes, temporary displaced thresholds, construction-related NOTAMs, and intersection departures introduce variability that must be accurately reflected in performance calculations. Any mismatch between assumed and actual conditions can significantly erode safety margins.
Case Study: LATAM Boeing 777-300ER Tail Strike, Milan (2024)
A particularly instructive example occurred on 9 July 2024 at Milan Malpensa Airport, when a LATAM Boeing 777-300ER on a long-haul flight to São Paulo sustained a severe tail strike during take-off from Runway 35L. The aircraft sustained substantial structural damage and was later classified as an accident.
The investigation revealed that the crew had used an incorrect gross take-off weight of 228.8 tonnes instead of the actual 328.4 tonnes—an underestimation of about 100 tonnes. The error occurred when the line training captain mentally subtracted expected taxi fuel from the displayed weight, yielding an incorrect figure. Crucially, this value was then communicated verbally in the cockpit and entered into both pilots’ EFBs for performance calculation using Boeing’s Onboard Performance Tool.
Because both devices received the same erroneous input, they produced identical thrust settings and V-speeds. This apparent consistency masked the underlying error and rendered standard cross-checks ineffective. The calculated rotation speed was more than 30 knots below the required speed for the actual aircraft weight.
During the take-off roll, the aircraft reached the erroneously computed rotation speed, prompting the crew to rotate prematurely. The aircraft pitched up rapidly but failed to generate sufficient lift for its actual weight, resulting in prolonged tail contact with the runway. The tail remained in contact for over 700 metres, causing extensive structural damage.
Despite the severity of the event, the crew managed the situation effectively after the incident by dumping fuel and returning safely to Milan. There were no injuries among the 398 occupants. However, the incident clearly illustrates how a single mental arithmetic error, combined with shared data entry and ineffective cross-checking, can breach multiple layers of defence.
Recurring Patterns in Recent Occurrences
The LATAM event is not an isolated case but part of a broader pattern observed in recent years. Several incidents have involved crews initiating take-off using performance data calculated for full-length runways, despite temporary reductions in available runway length. In other cases, aircraft have commenced take-off from intersection points without recalculating performance data, thereby reducing the usable runway length.
These occurrences are often linked to incomplete briefings, misinterpretation of runway markings, or failure to incorporate updated NOTAMS. In each case, the root cause is not the absence of procedures or tools, but a breakdown in verification processes and situational awareness.
Weight-related discrepancies have also been prominent, leading to incorrect V-speeds and thrust settings that increase the risk of tail strikes, runway excursions, or degraded climb performance. These events consistently follow a similar progression: a relatively minor input error is introduced, cross-checking is ineffective or compromised, and the resulting incorrect outputs are executed without challenge.
The Error Chain and Layered Defences
These incidents can be understood through the concept of an error chain, in which an initial mistake propagates through successive stages without detection. The failure typically begins with an incorrect input or assumption, followed by inadequate cross-verification, leading to incorrect performance outputs and, ultimately, unsafe execution during the take-off roll.
Aviation safety relies on multiple layers of defence to intercept such errors. These include organisational measures such as standard operating procedures and training; technical systems such as EFB validation logic and FMS safeguards; human factors such as disciplined challenge-and-response procedures; and operational elements such as thorough briefings and situational awareness.
When these layers function effectively, errors are detected and corrected before they compromise flight safety. However, when gaps align across these layers—as described in the Swiss Cheese model—errors can pass through all defences, leading to serious incidents or accidents.
Performance Margins and Operational Sensitivity
Take-off performance is highly sensitive to changes in aircraft weight, temperature, altitude, and runway conditions. As these factors increase, performance margins shrink, leaving less tolerance for error. High-weight, long-haul departures are particularly vulnerable because they operate closer to performance limits.
In such conditions, even minor inaccuracies in input data can significantly affect the required take-off distance, rotation speeds, and climb capability. This underscores the importance of accurate data entry and robust verification processes.
Strengthening Defences
Mitigating the risk of incorrect take-off performance data requires a combination of disciplined operational practices, technological enhancements, and organisational support. At the flight crew level, truly independent calculation and verification of performance data are essential. This requires avoiding verbal contamination of inputs and ensuring that each pilot conducts a separate, unbiased assessment before comparing results.
Reasonableness checks provide an additional layer of defence by prompting crews to assess whether computed values align with expected performance under the given conditions. Such checks can help identify anomalies that might otherwise go unnoticed.
Technological solutions are also evolving. Take-Off Performance Monitoring Systems (TOPMS) are being developed to compare actual aircraft acceleration during the take-off roll with predicted performance and to provide real-time alerts when deviations are detected. Enhanced EFB systems with improved validation logic and integration with aircraft and airport databases can further reduce the likelihood of input errors.
At the organisational level, scenario-based training that replicates real-world challenges—such as last-minute runway changes, intersection departures, and high-weight operations—can enhance crew preparedness. Flight Data Monitoring programmes can identify trends and detect anomalies.
Future Outlook
The industry is transitioning from: Error prevention → Error detection and recovery
Key developments include:
a) AI-driven anomaly detection
b) Real-time performance validation
c) Integration of aircraft and ground data systems
However, technology alone cannot eliminate the risk. The human-machine interface and operational discipline remain decisive.
Conclusion
Incorrect take-off performance data remains a persistent, systemic safety threat, not because of complexity but because of the failure to detect simple errors in time.
Recent incidents demonstrate that:
a) The hazard is not diminishing despite technological advances.
b) It is increasingly influenced by operational complexity and human factors.
The most effective mitigation is a multi-layered defence strategy:
a) Human vigilance
b) Technological safeguards
c) Organisational resilience
“Take-off performance errors are rarely unavoidable—they are almost always detectable before they become irreversible.”
Author: GR Mohan
