Garnet Ridgway says: Suggesting there are ‘no new accidents’ is a gross simplification and shows a lack of appreciation for the complexity of modern aerospace systems.
In an accident’s aftermath, all stakeholders are highly motivated to get answers; however, this is not the same as actually understanding what went wrong. For example, jumping to the conclusion ‘pilot error’ is a very easy and may superficially explain the sequence of events before an accident.
Consider the case of a military training aircraft flown by a novice pilot, which suffered an undercarriage failure immediately after touchdown. The investigation revealed the undercarriage lever was not set to the ‘locked’ detent; as far as the investigators were concerned, this was classic pilot error: case closed. However, that did not explain why the same issue occurred on two more occasions shortly afterward, despite changes to the training syllabus to emphasize the importance of locking the undercarriage before touchdown. Further investigation revealed that the touchdown ‘bump’ (in the hands of a student pilot) was sufficient to cause the undercarriage lever to leave the ‘locked’ detent of its own accord – a totally unacceptable design flaw for a training aircraft. The initial assessment of ‘pilot error’ was completely plausible, and completely wrong. As with the vast majority of accidents, it was in fact a totally new linkage between a hazard, a cause and an outcome.
A positive aspect of accident investigation is that the aerospace testing community is very good at learning from past mistakes – once accurately identified. Once a mitigation is in place and previous accident knowledge has permeated into the testing community’s thinking, it is extremely unlikely that it will happen again. This further supports the conjecture that there are, in fact, lots of ‘new’ accidents, as the causes of the ‘old’ are being iterated out of existence through mitigation strategies and measures.
It is human nature for previous accident knowledge to steer the investigation of new ones, but it can lead to tempting false conclusions. The challenge is accepting that not all accidents are ‘old’ accidents, and to avoid the pitfalls of assigning broad terms that give quick answers, but fall short of providing genuine understanding of causes.
Garnet Ridgway has a PhD from the UK’s University of Liverpool. He has designed cockpit instruments for Airbus and currently works for a leading UK-based aircraft test and evaluation organization
Sophie Robinson says: The history of accidents in aerospace testing is long and varied, stretching back to the earliest pioneers of aviation and touching the aircraft of the future.
Lt Thomas Selfridge was the first person killed in an aviation accident as a passenger during a demonstration of the Wright Flyer II in 1908. Aircraft accidents have become less frequent over time as both the technology available and our understanding of aviation safety have increased, and aerospace testing-related accidents in particular have become a newsworthy rarity; recent fatal incidents involving Virgin Galactic’s SpaceShipTwo and the prototype Finmeccanica AW609 tiltrotor made headlines around the world. In between these events, aerospace testing accidents have been numerous; surely, by now, we’ve seen everything there is to see, and everything that could go wrong, has?
Analysis of causal factors behind aircraft accidents reveals distinct trends in testing for both the civilian and military spheres. The human factor is the largest and most worrisome trend from recent years. Increasingly complex flight systems, aircrew fatigue and inconsistent training and regulation emerge as constant themes throughout air crash investigations. Evidence of repeated mistakes suggests that not all accidents in testing are new, and that changes to procedure and training are required to further reduce our accident rate. It is unsurprisingly that the same mistakes are repeatedly made within the testing community. We are, after all, trained in similar ways, from similar backgrounds, frequently working on similar projects, so it is not too surprising that similar things go wrong.
If we recognize mistakes are being made, what are we doing to prevent future accidents and further reduce test flying risks? Engineers commonly participate in the ‘learning from experience’ (LFE) process; both positive and negative issues and experiences during trials are documented. After completed tests, this knowledge is fed to the wider engineering community so that issues can either be mitigated or prevented in future. A positive, pro-active safety culture is driving a reduction in testing accidents. As this rate iterates toward zero, a time will come when testing accidents are confined to the history books.
Sophie Robinson works at the front line of aerospace testing as a rotary-wing performance and flying qualities engineer for a leading UK-based aircraft test organization. She also holds a PhD in aerospace engineering from the University of Liverpool