Learning from experience? How? Some research issues in industrial risk management

Jens Rasmussen

Research output: Chapter in Book/Report/Conference proceedingBook chapterResearchpeer-review


Modern large-scale industrial systems require special precautions for safe operation and systematic risk analysis is frequently used during system design. The paper reviews a number of problems presently found in the use of risk analysis as a basis for effective risk management.There is a need for more explicit formulation of the preconditions of safe operation and for better communication to the operating organization. Operation in a competitive environment requiresendless adaptation and optimization by management and, consequently, the preconditions for safe operation are likely to be violated. This is even the case in systems designed according to the defence-in-depth principle and several major accidents have happened during the latest years.How do we improve this situation? The paper reviews a number of problems found in analysis of accidents and in generalization of the lessons learned into effective safety improvements. The difficulties in identifying root causes and of finding the factors in a complex organization which are sensitive to improvement are discussed in some detail. The questions are: Is the repeated experience of major accidents a manifestation of the accepted risk in modern technology? Is it the result of inadequate methods for analysis? Or is it the result of ineffective methods to generalize and to identify factors sensitive to improvement?
Original languageEnglish
Title of host publicationLes facteurs humains de la fiabilite dans les systemes complexes
EditorsJ. Leplat, G. de Terssac
Number of pages18
Place of PublicationMarseille
PublisherEdition Octares
Publication date1990
ISBN (Print)2-906-769-03-7
Publication statusPublished - 1990
Externally publishedYes


Dive into the research topics of 'Learning from experience? How? Some research issues in industrial risk management'. Together they form a unique fingerprint.

Cite this