It has been rather widely argued and demonstrated that the lessons to be learned when developed during investigations of prior accidents are not being implemented very well or very promptly, and sometimes not at all. When I started probing present investigation practices to try to understand why this was happening, serious impediments to the process of learning, reporting, disseminating and, particularly, implementing lessons produced by investigations began to surface. Each impediment I found rapidly led me to others - and also to new insights about the more fundamental problems and challenges leading to them. I have reported these in a series of papers and presentations since 2007.
About a year ago, I pulled together all these papers, presentations and tutorials, and posted them on my web site. along with a list of the new insights gained during preparation of each During the preparation, it became apparent to me that current investigation practices are designed to satisfy investigators’ needs, imposed on them by regulatory or organizational demands to find causes and make prevention recommendations. But the actual changes in existing operation to prevent or reduce risks of future accident can only be made by system operators - the end users - by changing from what people, objects or energies are doing now, to new behaviors and behavior patterns.
To do that today, operators must find what investigators produced and then translate investigators’work products into actionable behavioral information so they can determine if it is relevant, and how they might adapt it to modify what their people, objects or energies do in the future. Investigations are delivering causes and recommendations, which are not directly actionable behavioral information for the end users to implement . The implication was inescapable: Investigations are serving the Wrong PARTIES! Rather than satisfying the end users’ needs for actionable behavioral information, investigators are satisfying investigators’ needs with causes and recommendations. Additionally, present investigation work products make it impossible to measure investigations’ contributions to the prevention or reduction of risks achieved.
The disparity has led me to challenge the traditional paradigm where “cause” and “ safety recommendations” drive investigations. If the main objective of the investigation process is prevention or the reduction of risks, then its highest priority must be serving the needs of the end user/operator community who must make that happen. The way to accomplish that is to change accident investigators’ mandate, from the current causation-recommendation determination mandate, to a new mandate to deliver timely, actionable behavioral information serving end users’ real world needs.
A replacement paradigm for investigation processes is needed to properly serve end users’ needs with actionable behavioral information they can use directly to change behaviors in a timely, effective, efficient and verifiable manner. The papers at www.iprr.org/research/llprojcontents.html, offer an alternative, within a lessons learning system framework, investigation process design criteria, data transformation, analysis and reporting details and metrics.
