Management of Precursors

National Academy Of Sciences, Washington DC

Investigators

Abstract

Many accidents result from a series of events, "an accident chain", with events occurring in just the right, or more aptly, just the wrong way. The events that precede and lead up to the accident are termed "precursors," and they represent a rich source of information to project extreme events, assess risk exposure, and determine appropriate risk mitigation measures. Because precursor events can be used to understand and mitigate future risk exposure, industries have developed analytical techniques and management approaches for detecting, modeling, and acting on precursor signals. These industry methods have been developed largely in isolation from each other and pollination of methods across industries has been limited. This study will provide a platform for sharing accident precursor methodologies across industries, and will scope future research directions and challenges in precursor management. The study, to be conducted by a committee of experts from the risk management and engineering communities, will analyze precursor methodologies in the airline, chemical, health care and nuclear industries, among others. These industries face the common challenge of mitigating against low-probability, high-impact events, yet are dissimilar enough such that approaches have evolved with limited cross-industry sharing. The study will include a review of the literature, expert testimony, two workshops, committee deliberations and other fact-finding activities. The study will culminate in a National Academy of Engineering consensus report that incorporates committee findings and workshop proceedings. The report will: 1. Present a framework for understanding precursor events in the context accident sequences. This framework will have cross-industry applicability and be consistent with current industry practices. 2. Highlight research findings in modeling precursors and present future research directions and challenges. 3. Examine the success of mandatory and voluntary reporting systems in capturing precursor data. The report will present the factors, policies, and methods that can affect the performance of these systems. 4. Present the successes and challenges of designing warning systems for the automated detection and diagnosis of precursor events. The report will provide industry practitioners, researchers in risk management and safety science, policy makers, and public-sector agencies with guidance on using precursors for risk mitigation. It is envisioned that the report will benefit a broad range of industries and disciplines in mitigating against events for which there may be little, if any, direct experience.