Practical theories of safety and risk

I think it [risk] depends how many defences are in place. Have you lost one out of ten or one out of one How well is it covered by other systems and procedures That is partly it. Also, had you lost the next defence, was it going to be really terrible Something that has the potential to cause harm is serious. (Si3-3) 

When it is down to a single piece of kit or just the crew, experience has proven that it is not a good defence at all in a lot of cases. 

Organisational safety requires having systems and work routines in place that are able to respond to failures - and making sure that those systems stay in place at all times and are activated when they may be needed. For investigators, their ideal of safety depends on ensuring there is a persistent capacity to respond to disruptions and to call on latent systems and work routines that are lying dormant, ready to be activated if needed. 

Taking the loss of comms event - the potential is a mid-air [collision] but... you ve still got ATC radar, they ve got alerts, you ve got TCAS [traffic warning system], you re looking out the window, you ve got HF [radio], you lose your HF you can come back. (Si6-5) 

In any situation, organisational safety depends on multiple and diverse capabilities to deal with unintended and unexpected disruptions social and technical redundancy. Safety also depends on maintaining excess defences and capacity in reserve, beyond what is actually called upon to catch a failure. Investigators are interested to see how well a particular 

---