Models for mission critical systems

The Smith Institute helped Wittenstein aerospace and simulation to optimise maintenance strategies.

The problem

The effective maintenance of mission and safety critical systems aims to minimise the costs of maintenance while meeting specified limits on the likelihood and consequences of system failure, and to do so in a way that also responds to the operational profile and environment of a given system.

Wittenstein are constantly seeking to improve their testing procedures and the management of product reliability. They engaged with the Institute to develop a framework that will support testing of a system during periodic maintenance, enable optimisation of the system replacement strategy on the basis of test results, and provide proof to customers of an acceptable level of reliability of the system in the subsequent period of operation. During the work we developed framework for testing an electro-mechanical system and determining from test results and from the system’s past test history whether it should be (1) replaced immediately; or (2) operated for one period and then replaced; or (3) operated for one period and then retested.

The solution

The Smith Institute worked with Wittenstein aerospace & simulation limited to implement the latest mathematical thinking to cost-effective maintenance that also meets operational and safety requirements of airborne systems.

The mathematical framework includes a system model whose parameters are estimated from test results using Bayesian methods. Underlying processes of wear are estimated by Kalman filtering, and they are used to determine confidence levels for the system performance being within functionally acceptable limits in subsequent maintenance periods. A decision-support model, which minimises the costs of maintenance while also meeting limits on the probability of system failure, indicates the actions that should be planned or taken during maintenance periods.

The benefit

The framework has provided Wittenstein with insights into methods for system fault diagnosis and prognosis, for probability and cost-based decision-making and for the optimisation of maintenance strategies.

The Smith Institute helped us to develop a precise problem definition and to formulate a sound mathematical basis for our future work.

Dr David CowlingManaging Director of Wittenstein Aerospace & Simulation LTD