The worst crash in aviation history between two 747's in Tenerife in 1974 was a result of pilot impatience coupled with bad weather and unclear communication between the pilot and co-pilot.
When a small commuter plane crashed on take-off in Charlotte, North Carolina it was a result of inaccurate maintenance coupled with outdated weight computation methods.
The crash of an Avianca plane into Long Island after it ran out of fuel was a result of one of the pilots not having good English coupled with the second pilot not clarifying the emergency nature of the fuel.
With all these accidents any one of the issues in isolation could have been dealt with, but add them all together and you get to a situation where an accident is bound to occur
This got me thinking about processes and single points of failure.
Will a process on the whole fail as a result of a failure at one point in the process? Sometimes the answer is 'Yes' (In my last post about the use of procedures by NASA, missing a single step in a procedure could quite easily jepoardise the whole shuttle flight), but in many situations a single failure will not be catastrophic. An earlier post I wrote referred to bad process design in airline web-sites. In the example quoted, a bug meant that addresses with punctuation in them (such as the period in 'St. Leonards' would be rejected) but this could be bypassed by removing the offending punctuation and the booking could continue. But remove the punctuation, omit an expiry date on a credit card payment and put an incorrect email in the relevant fields and the booking is doomed to failure.
How many of your processes have a single point of failure? Should they have one, or more than one? Or none?
Reminder: 'The Perfect Process Project Second Edition' is now available. Don't miss the chance to get this valuable insight into how to make business processes work for you. Click this link and follow the instructions to get this book.
All information is Copyright (C) G Comerford
See related info below