How do we improve product or system reliability performance through good design?

At first sight this question seems quite vague and open-ended, but applying the ideas of reliability engineering enables us to focus on a few key factors that can help to provide an answer.

Asking the right question is usually the key to finding the right answer, so perhaps we should start by asking:

What type of plant, system, equipment, etc. are we considering? Is it

1. mission based and cannot be repaired during the mission?
   OR
2. process based and capable of being repaired or maintained during the process?

Mission based
If the answer to the above question is a) we must design the system to have a high probability of achieving its mission. In other words it must have a high reliability. This can be achieved by a combination of the following:

1. Using high reliability components, i.e. using better quality components
2. Incorporating redundancy, i.e. duplicating functional elements of the system
3. Incorporating diversity, i.e. using components from different manufacturers and/or components that are diverse in design or operation

The emphasis given to each of these three options will in practice be determined by criteria other than reliability such as cost, size and weight of the final design.

Process based
Alternatively, if the answer to the above question is b), we must design the system to have a high average output or throughput. In other words it must have a high availability. This can be achieved by maximising the reliability of the system relative to its average downtime. Creation of a reliable design in this instance implies not only improving the reliability of the system in the same ways we considered above for a mission based system, but also minimising the average downtime. Options for minimising downtime include the following:

1. Improving maintainability (i.e. reduced maintenance and repair times) by
   – quick deployment of maintenance personnel when failure occurs
   – efficient diagnosis of the cause of failure
   – improved component access and reinstatement for repair or replacement activities
   – provision of effective tools, training and procedures for maintenance personnel
   – ready access to spares for replaceable components
   – use of condition monitoring to provide early warning of failures so that outages can
   be scheduled to minimise their impact on output.
2. Incorporating by-passes and other features to allow flexible or improvised operation when necessary (this is often overlooked in the design phase). However risks that may be introduced through failure or mis-operation of such features should be investigated and balanced against the benefits.
3. Incorporating redundancy and buffers (particularly storage) in the design. Although normally used to improve reliability, they can also effectively improve availability by reducing or avoiding downtime during routine maintenance (this is a benefit that is also often overlooked in the design stage).

Although engineering systems can be widely different in design and operation, they are either “mission based” or “process based” and the ideas above can be applied generally. However, for both mission and process based systems the emphasis given to the various options for improving reliability or availability will in practice be influenced by other considerations such as cost.