The Neglected Role of Information Measurement in Failure Prevention and System Availability

摘要

Information measurement, with an emphasis on its relationships with competent decision-making, is being investigated to seek out the proper roles for employment of information measurement concepts in failure prevention and system availability. While the science behind information measurement can be exceedingly sophisticated, and at times difficult, it is suggested that assessing current failure prevention/system availability practice in light of what appears to be of interest for improvements, can itself be straightforward. Figure 1.0 outlines the approach selected for starting the process, primarily through the introduction of and/or review of concepts with the firm hope that useful applications and opportunities are "unmasked". The computational and communication systems impact on our daily lives has been profound. It is strongly suggested that, not only is there more to come, but also that 'hope' can be offered when more and proper respect is paid to the fundamental concepts associated with Information Measurement, particularly as it impacts decision-making. Figure 2.0 provides an estimate that, while in today's world the people who live along the Hudson River have 400 times the income of those who live along the Brazilian/Venezuela border, they have 30 million times as many items to choose from when "spending their money". The data noted on Figure 2.0, from a small piece in the referenced Scientific American (reference 1.0) under the title 'Evonomics', could be considered a number of ways, ranging from an assessment that we (surrogate Manhattan people) are already "pretty good" at information processing, though simply recognizing that a "data avalanche" is underway. It is suggested that the shopper's decision-making conundrum parallels what we all face with the "data intensive" scenarios created with our 'essentially undisciplined' love of computational and communication "data" technology. Those interested in failure prevention and system availability are no exception. Information is 'sine-qua-non' for intelligent decision-making. Hence, in an effort to investigate the applicability of 'information measurement' itself, it is suggested that examining the basic elements for decision-making provides an acceptable 'laboratory'. Going to the next level, it is further suggested (Figure 3.0) that decision-making inputs can be 'essentially captured' by considering three elements: i) facts; ii) conditional/Bayesian features; and iii) quantitative characterizations, including probability density function (pdf) selections. If there is an 'umbrella' concept for considering information for supporting effective decision-making, it is suggested that it would be the need to "consider all of the information". This is not always the case. As noted in Figure 3.0, it is suggested that one or more of the applicable input information elements for decision-making are "too often" not processed and/or "screened" to extract their information content. An assessment of each of the Figure 3.0 suggested input information elements will be briefly considered as they relate to current practice.