High reliability organizations are very biased towards highly-reliable work processes. One for an airline pilot is illustrated in Figure 7-11. The pilot has an error rate of maybe one in 1,000, as an example.
If you lay out a mistake that the pilot made that results in loss of an aircraft, you could say, “We lost the plane because there was an error. There was an error because the pilot missed step two of the checklist.” If you put a checklist in place and pilots have an error rate of one in 1,000, then maybe one in 1,000 or so times a pilot could miss step two.
If you use a process where the copilot also checks the same checklist, the only way you could have the error is if both the pilot and copilot miss step two. These errors then become one in 1,000 times one in 1,000, which is one in a million. That process is illustrated in Figure 7-12.