The article linked in the OP, and other views, are based on the official investigation report. But the limitations in scope and reporting of accidents (Annex 13 – factual requirement) might not be suited to identifying safety improvements for an already very safe industry, particularly where accidents more often involve the interactions of many contributing factors.
Formal reports identify ‘what’ happened and to a limited extent ‘why’; the why is increasingly restricted by the variability of human involvement.
An alternative is to consider ‘how’ the accident could have occurred, not to establish any factual clarity, but with reasoned speculation identify contributory (latent) factors in the specific accident and similar events as a basis for proactive safety improvement.
A reactive safety process is guided by the report, but all that might achieve is avoiding the accident which has just occurred. Similarly a speculative approach seeking a single solution can be subject to hindsight bias, as reflected in many discussions with polarised views on the human vs automation.
However, hindsight, without bias of cause or blame could be used for learning, increasing industry’s knowledge of operations and of ourselves, and from this identify areas for improvement.
What could be learnt from this accident; what have we learnt and how – without the biases?
Learning is often based on enquiry or error, where concepts opposed to facts are more valued. We learn from history, we learn from others, and from activities not directly related to the proximity of an event; i.e.
- The contribution from regulation, the airworthiness process and how humans are viewed - a threat or an asset.
- The contribution of the training process, how are regulations interpreted by operators, implemented and presented to the crew; including formulation of drills and checklists.
- The contribution of behavioural shaping in normal operation, how crews gain experience, assess situations, and make decisions.
Instead of debating strongly held, often biased and factually unprovable views, we should consider what we, individually and collectively, could learn from accidents.
The relative merits of factors, separately and in combination would still be debateable, but at least there might be a greater range of views to consider for implementation.
The choice of implementing ‘concepts’ (if that is possible) would not depend on regulation, or the factual limitations of past events; the choice would be ours, we choose what to learn, what to consider, what might be of value, what if …
From these thoughts we might then be able to turn the hindsight biases of media interpretation and individual views into the much needed foresight to maintain and hopefully improve safety.