Read more
About the author
Robert Taylor is an engineer with 55 years of experience in research, practical design and of international consulting in the field of risk analysis. Starting at the Theoretical Physics Department at the UKAEA Harwell research laboratory, he moved to the Risø National Laboratory in Denmark where he worked in development and validation of risk analysis procedures for major hazards plants, prevention of design error in nuclear power and in aerospace systems, in software safety analysis and working together with Pr. Jens Rasmussen, on the development of validated methods for human error analysis.
Together with his wife Hanne, in 1987 he formed a company to undertake international consulting, with projects on every continent except Antarctica.
Major achievements were development of automated methods for hazard and operability analysis and software fault tree analysis, development of an extended evidence based human error probability database, performance of follow up studies of the accuracy of risk assessments compared with actual experience for over 100 plants, and development of a series of methods for identification of design error, one of which became an international standard.
Summary
This volume describes method which can be used to prevent design errors from arising. The principles and approaches to error avoidance are described including improvements of the design process, improvements in knowledge availability, and improvements in communication.
Computer aided design has become a major tool in preventing some kinds of error. Especially modern 3D object based design allows greater understanding, and investigation of detailed feature of designs can be odeled. Computer aided design has also introduced new ways for committing them.
Simulation, including the use of digital twins, has become a major tool in checking for error in some disciplines, and examples are given of its use. The problems in achieving model fidelity, coverage of a full range of failure possibilities, and selection of challenges to test the simulation moded, remain as issues.
The human machine interface is also an important area for errors and weaknesses, and current guidance on HMI design is described. This is largely based on ergonomic and interface design requirements for nuclear reactors, but application to other industries is described.
There is a range of design principles, such as avoidance of single point failures, which are described and illustrated in the volume. All of these have been tried and tested in practical projects.
With the prevalence of errors in initial design, design success is often dependent on effective design review. Methods for design review are described
Change is generally necessary for any design, bot as designs are integrated, during construction or production, and in actual operation. Changes must be managed, otherwise they themselves can be a cause of failure. The importance of, and methods for, management of change are described.
