Turning Aviation Data into Better Training

Evidence-based training, or EBT, promises to make crew training relevant by using operational and performance data to identify what needs to be trained. The challenge is turning that evidence into effective training, and demonstrating that it improves safety,

The principle behind EBT is simple: training should follow the data, not the other way around. Rather than drilling crews on a fixed set of scenarios year after year, operators look at what is happening in their operation – incidents, hazard reports, flight data monitoring, real assessment outcomes – and use that evidence to shape what gets trained, and how.

In practice, turning that principle into a working programme is far from straightforward. Data must be sourced, weighted, and interpreted; generic competency frameworks must be adapted to an operator’s specific fleet, network, and culture; and, perhaps hardest of all, someone must prove the whole system is improving safety rather than just producing the appearance of rigour.

Turning evidence into training

EBT draws on data from both the wider aviation industry and operator-specific sources, including hazard and accident reports, flight data monitoring (FDM), accident and incident investigations, manufacturer information, and data collected during EBT assessment and training sessions, according to EASA.

“We do not prescribe a fixed weighting between these sources. The operator, through its management system, safety, flight operations, and training functions, analyses the available data, customises the EBT programme, and contextualises the elements of the example scenarios for presentation by appropriately qualified EBT instructors. This requires operational expertise, safety data analysis, and expertise in EBT program design,” an EASA spokesperson says.

Raghoonundun Gunputh, Senior Manager Safety Training Policies, Standards and Legal Requirements at Lufthansa Aviation Training, points to cabin crew safety training as an example. The International Civil Aviation Organization’s (ICAO) Cabin Crew Safety Training Manual (Doc. 10002) provides a generic competency framework that operators can use as a foundation for competency-based training and assessment. However, that framework needs to be adapted to reflect the particular training exercise and operational environment.

“While the generic framework provides a valid foundation, some observable behaviours may not be relevant to every simulated exercise or operational environment. Therefore,

developing an ‘Adapted Competency Model’, as specified by ICAO Doc. 10002, supported by customised performance criteria and evaluation metrics, is essential,” he says. “A customised framework allows instructors to focus observations on behaviours directly relevant to training objectives, resulting in more accurate assessments, more meaningful performance data, and ultimately, more effective training outcomes.”

An EBT programme becomes specific to an operator by adapting assessment and training to reflect how that airline operates, including the fleet, network, procedures, operating environment, flight personnel, training data, and operational safety information, as well as what its training and safety data reveal.

“For baseline EBT (the most basic approach), a mature management system and a fully developed feedback process are particularly important for monitoring training system performance and continuously improving the program,” the EASA spokesperson says.

“Necessary conditions include effective cooperation between safety, flight operations, and training functions, reliable reporting supported by a positive safety culture, robust instructor standardisation, and adequate safeguards against the misuse of EBT data. These safeguards allow the data to be used to identify collective and individual training needs and develop flight crew competencies.”

Making sense of the data

For some areas of aviation training, however, the systems and datasets needed to make fully evidence-based decisions are still developing.

Gunputh notes that competency-based training and assessment principles are still at an early stage in cabin crew training.

“It will be several years before the industry has sufficiently mature systems and meaningful datasets capable of supporting evidence-based decision-making. In the meantime, the most effective approach is to systematically collect and analyse training data using advanced analytical tools that translate observable behaviours (OB) into meaningful performance insights,” he says.

“This analysis should be based on clearly defined taxonomies that allow operators to identify strengths, weaknesses, and performance trends across different competencies. Rather than simply recording performance results, these taxonomies should help determine whether there is a reduction in safety margin and, if so, support the implementation of proactive corrective actions, such as targeted training interventions or skill enhancements.”

According to EASA, the effectiveness of EBT is monitored through the operator training system’s performance evaluation process.

“Relevant indicators include proficiency levels, observable behaviours, proficiency/non-proficiency outcomes, recurring errors, and instructor agreement. The effects of a training intervention may become visible in the short term through changes in specific operational behaviours or safety performance indicators, particularly when the intervention addresses a clearly defined issue and there is sufficient exposure,” the spokesperson says.

From observation to action

Collecting data is only useful if operators can identify what lies behind the performance it reveals.

Identifying ‘poor performance in workload management,’ for example, is too general to provide meaningful guidance for training improvement, says Gunputh.

“To effectively address performance deficiencies, operators must understand their root causes. This can be achieved by categorising observations into more specific taxonomies, such as ineffective task delegation, reduced situational awareness, delayed decision-making, or cognitive blockage caused by surprise. This grouping would allow for more targeted and effective training solutions,” he says.

Developing and continuously improving skills across the entire pilot population will generally require a longer period, as skills must be practiced, assessed, and consolidated over the course of subsequent EBT modules, according to EASA.

“We do not prescribe a fixed timeframe; conclusions should be based on sufficient exposure, several complementary indicators, and sustained trends, rather than single events or short-term fluctuations. Pilot feedback can also provide a preliminary qualitative indication of effectiveness, as EBT is often perceived as more relevant to day-to-day operations due to realistic, contextualised scenarios and skill development,” the spokesperson says.

Better evidence, better training

EBT is as much a discipline of interpretation as it is a data-collection exercise. EASA is deliberately light-touch on prescribing weightings or timeframes, leaving operators to build the safety-culture infrastructure, needed to make sense of their own data responsibly.

Raw observations mean little on their own, it is only when an observation is broken down into specific, well-defined categories that it becomes something training can target.

In some areas, competency-based training is not yet mature enough, nor the datasets large enough, to fully close the loop between training intervention and measurable

safety outcome. But the direction of travel is clear: better taxonomies, more disciplined data analysis, and patient, sustained observation over single-event snapshots.

As these systems mature, evidence-based training should move closer to fulfilling its original promise: training that improves tomorrow’s safety margins.