'It works in practice, but not in theory'
A major problem with resilience depends on how it is defined; this is where training discussion starts from - as questioned in post #1.
The EASA view, and other individual training initiatives consider Resilience as something the human will have, thus should be trained. As a continuing evolution of individual HF, team CRM, and thence resilient human performance for uncertain, complex situations.
Alternatively most of the academic views consider Resilience as a capacity of a system.
The distinction is identified by the term Resilience Engineering - how to construct a safety management system for a complex and uncertain future. This requires a change in safety thinking, revised views of safety management, new vs old view of safety, SI and SII, which as yet is not happening, and something where regulators have to lead - adapt their thoughts about safety management.
Arguably, individual and team training is approaching a limit of effectiveness; flatlining safety statistics, difficulty in measuring results based on failure in a safe industry - fewer accidents. Further training, if at all, has to evolve from how to avoid failing, to how to succeed in unforeseeable situations. This might focus on aspects of individual surprise and risk management, not seeking to change behaviour, but manage it; also consider how natural human resilence can complement a system as an adaptive component.
The need of a new paradigm for safety management (Resilience Engineering) is well represented in several papers, but again arguably, not yet adapted possible because of the current high level of safety, no need to change - but we should not have to wait for the surprising events - proactive safety management.
AF 447 discussions are a good example of the differing viewpoints.
- Errant human behaviour, old safety, SI, more training, … (a confusing view of personal resilience)
- A systems view; how-come the previous 20 or so events did not result in an accident, even with some similar human behaviour initially. What was learnt about human performance from these events?
The conclusion was that the ADS design in rare situations was not as assumed by certification, and thus required to be changed. The focus was technical.
Alternatively a joint tech, human, situational system view might have provided better interim safety intervention; crews did not respond as assumed by the checklist, but did recover, changed activity. e.g. simulator training for unreliable airspeed flight actions, vs the awareness of do nothing - read the checklist and reassess the situation (not economic use of a simulator).
Last edited by alf5071h; 18th Mar 2024 at 14:05.
Reason: Sp