Air safety: hazards, risk and the barriers between them
Hazard, risk and the two-axis test
A hazard is a source of harm: a busy road under the take-off point, a 15 kt gust, a partially charged battery, a helicopter route. Risk is the combination of the likelihood that the hazard causes harm and the severity of the outcome. Questions frequently test whether you assess risk on both axes - a low-probability collision with a manned aircraft is still unacceptable because the severity is catastrophic.
Barriers and the Swiss cheese model
Barriers sit between hazards and harm: geofencing, the return-to-home failsafe, an observer, a pre-flight briefing, a battery log. The Swiss cheese model teaches that each barrier has holes and that holes can line up. Worse, barriers can share a common-mode failure - a geofence database that is out of date and a pilot who never checks NOTAMs give you two barriers that fail together. Never treat a proposed barrier as effective unless you have verified it works.
Reporting and just culture
Occurrences such as fly-aways, near misses with manned aircraft and injuries must be reported so that others can learn. A just culture separates honest error from recklessness; it is why people report. Concealment turns one accident into a pattern of accidents.
Trap: 'the aircraft has a failsafe, so the risk is nil.' Untested or misunderstood automation breeds complacency and is worse than no automation at all.
Rule of thumb - the 30-second stop: before every flight ask three questions: what could go wrong here, what exactly would I do if it did, and what would make me not fly? If you cannot answer all three, you are not ready.