Karl Weick and Kathleen Sutcliffe have written an eloquent and practical guide to reliability and safety that emphasizes the managerial point of view, but is also equally helpful to researchers or safety professionals. Perhaps the greatest thing the authors are able to accomplish in this book is in emphasizing the conscious mindfulness required in critical situations, and in distinguishing in observable and real-world ways the specific components of mindfulness as seen in safety-conscious High Reliability Organizations (HROs).
The authors distill the essence of reliability (and safety) into five essential qualities: preoccupation with failure, reluctance to simplify, sensitivity to operations, commitment to resilience, and deference to expertise. As a long time safety professional (with experience largely in the aviation and chemical processing industries) I couldn't agree more with the authors after reading the text associated with these five qualities. I have found that especially in larger organizations that deference to expertise is perhaps the most difficult of the five traits to be accepted in the workplace, as generally rank or seniority tend to be deferred to, particularly in a crisis. The airline industry has come a long way with the different iterations of Crew Resource Management (CRM), and of all (often unstated) the reasons that CRM has succeeded I think that deference to expertise is the single most important.
I like the concept of realistic audits the authors promote, and particularly enjoyed the insight regarding the vulnerability of Singapore to Japanese attack as it came to be understood by Winston Churchill, who had a penchant for realistic self-appraisal, to wit: "I ought to have known. My advisors ought to have known and I ought to have been told, and I ought to have asked." The point is that we frequently believe what we want to believe, not because we are intellectually dishonest, but because of the human tendency to seek out information that confirms our views, and not to seek out disconfirming information. A mark of a truly reliable and safe organization (examples include airline operations, nuclear power plants, aircraft carriers, etc.) is seeking out information which points toward problem areas, rather than viewing successes as being demonstrative of the quality of institutional planning and procedures. The example concerning the Moura mine disaster on p.135 makes the point quite eloquently: "HROs assume that the system is endangered until there is conclusive proof that it is not." There could be no better single- sentence summary of the book.
There are many more interesting observations in the book, the most enlightening of which can be found in chapter five. I particularly enjoyed the discussion of a "learning culture" beginning on p. 136, and find that one of their most salient observations is also one of mine from years of industry experience, that being the concept of "learned helplessness." When people attempt to bring safety issues to the fore but are quickly dismissed, they learn to keep to themselves. This is a major problem, especially in large organizations, and the advice proffered by the authors is both sound and cogent. I was absolutely delighted to see the long-overdue examination (p.140) of "de minimus error" in which context is examined as it relates to seemingly unconnected small events. In this situation, people frequently seek out separate, small reasons for each deviation, ignoring the accumulating evidence that there is actually one large problem responsible for all the disparate events. Though the authors did not note it as an example, people familiar with the Apollo 13 accident will no doubt realize how the controllers had to fight off this kind of error willfully. (I think that Apollo-era NASA was an excellent example of an HRO.)
There are many more issues that Weick and Sutcliffe bring to the forefront in this book, from intelligent rule-making, to flexibility of response. My advice to any manager or safety professional is to put this book at the top of your reading list. It is easy to read, easy to digest, comprehensive in scope, yet universally applicable across industries. Even if you are not involved in an industry like nuclear power or aviation where large issues of life and death are literally in your domain, this is still mandatory reading. Any business can learn for the examples cited (which range from a merger-induced railroad meltdown at the "bad" end of the reliability scale, to nuclear-powered aircraft carrier operations at the "good" end of the reliability scale.)
I highly recommend this book to managers, safety professionals, researchers, and anyone else interested in becoming more informed about reliability and organizational safety.