REMINDER: The General Applications Section of the Royal
Statistical Society is hosting a half-day meeting on
RELIABILITY AND MAINTENANCE
at the YMCA (next to the RSS offices) in Errol Street, London, on
TUESDAY 12TH DECEMBER 2000
The programme is detailed below, followed by titles and abstracts.
Admission is free and everybody is welcome to attend, without
prior notification.
PROGRAMME
2.00pm JOHN ANDREWS
2.45pm FRANK COOLEN
3.30pm Refreshments at the RSS
4.00pm ROSE BAKER
4.45pm JEFF JONES
5.30pm End
Advances in Fault Tree Analysis
Dr John Andrews (Loughborough University)
Fault Tree Analysis is the most common technique to predict failure
probabilities and frequencies for systems. Many system assessments
are conducted at a basic level and advanced features are less common,
such as Initiator and Enabler Theory. This presentation introduces basic
and advanced capabilities by means of an example system. The Binary
Decision Diagram will also be discussed.
Bayesian Graphical Modelling for the Software Test-retest Cycle
Dr Frank Coolen (University of Durham)
A method to support software testing, using Bayesian graphical models,
has been developed in close collaboration with an industrial partner. An
overview of the method will be given, with special attention to test
design, which is one of the most important and most complex issues of
software testing. This is based on joint work with David Wooff and
Michael Goldstein.
A Methodology for Sensitivity Analysis of Maintenance and Reliability
Models
Dr Rose Baker (University of Salford)
A simple methodology is presented for sensitivity analysis of models
that have been fitted to data by statistical methods. Such analysis is a
useful decision-support tool. The methodology is illustrated with
examples. A simulation study using a simple model of maintenance is
used to exemplify the accuracy of the method as a function of sample
size.
The Use of Discrete-event Simulation to Model the Achievement of
Maintenance-free Operating Periods for Aerospace Systems
Mr Jeff Jones (University of Warwick)
This presentation describes a discrete-event simulation model that
allows a system designer to investigate the means and feasibility of
meeting an operating period requirement. The model takes into account
system reliability, proposed mission tasking and maintenance support.
The basic output is the probability of achieving an operating period of a
desired length, although other outputs are possible.
|