Capital Equipment Support: Management of Repairables

Introduction

Large and complex items of capital equipment such as aircraft, locomotives, ships and weapon systems can only justify their considerable investment costs whilst they are working or available for work.

The cost of supporting a complex item of capital equipment throughout its life often exceeds the original purchase price by a considerable margin. Determining a strategy for the provision of support is therefore very important. This article is a very brief summary of some aspects of repairables strategy and management.

Objectives

Fig 1 illustrates the classic trade-off between service level and cost. Any performance to the left of the 'optimal trade-off line' represents waste - better service can be achieved for the same cost or the same service for less cost. Every organisation should therefore try to get as near as possible to the ideal represented by this line. Determining which point on the line to aim for is a matter of policy rather than economics.

Service level may be defined in many ways. In the context of repairables, most definitions relate to the availability of the parent equipment, not the individual components of which it is comprised.

Cost is usually defined to be the (discounted) cost of supporting the equipment throughout its life. This includes the initial stock of repairables, the consumables, materials, labour, training, documentation, support and test equipment.

It is increasingly recognised that the ultimate target should be to minimise total Life Cycle Cost (LCC) incorporating not only support costs, but also the original purchase cost.

This implies that consideration of the downstream costs of providing support should be an integral part of the design process - from initial concept to production planning This is the concept of Integrated Logistics Support (ILS) now underlying much US defence procurement thinking.

Methods and Locations

Repair by replacement is potentially recursive: it generates a failed component which can be viewed as a mini-equipment for which the same economic considerations apply. The structure generated by these recursive relationships is illustrated in Fig 2. The branches terminates when a repair no longer generates any defective components which can be cost-effectively repaired.

Determining repair policy is therefore a matter of designing a tree. A typical tree is shown in Fig 3.

Each branching point is a location at which a component of a specific type is repaired. Each branch is either a repair loop or a 'scrap and replace' terminator. The design decision must consider the impact on the overall service level and the overall cost of each branch and each branching point.

The problem is potentially very complex. Each additional repair loop incurs such overheads as: labour, training, test rigs and jigs, and an investment in a pool of serviceable repairables (the repair pool). Reducing the number of repair loops usually reduces the overall service level as well as these costs. For geographic reasons it is often necessary to undertake the same type of repair in more than one location.

In all but the most trivial cases therefore, obtaining reasonable answers is a job for computer-based mathematical models.

Repair Loop Configuration

For each repair loop and each repairable component, there must be a repairables pool (a number of components surplus to the immediate equipment requirements) which exist to ensure that when a serviceable item is required, there is a good chance that one will be available.

Scheduled Servicing and Failure Processes

If a component is prone to early failure (as is the case for many electronic items) then it may be appropriate to "burn it in" before selling it. If a component's failure rate increases with age, then scheduled services may be a good idea. If the failure of a component has safety implications then scheduled servicing is even more relevant.

Determining what the best interval between scheduled services should be is yet another trade-off.

The more frequent the service, the greater probability of wasting effort, but the lower the likelihood that premature failure will reduce the predictability of service schedules.

The relevance of scheduled servicing is crucially dependent on the shape of the "age-specific failure rate curve". If it increases sharply at some point then planned servicing is more likely to be effective.

The "bath-tub" curve illustrated in Fig 4 is the much-quoted, but actually quite rare type of age-specific failure rate which exhibits both "infant mortality" and ageing.

Management

Even if the design is good, day-to-day management of repair loops is essential.

Serviceable repairables are usually controlled quite effectively. By contrast, defective repairables are sometimes ignored until it is too late. When an out-of-stock situation occurs the number and location of defective items is not always accurately known. When they are found, they are often found at the scene of their failure.

Sometimes quite large queues of defective repairables build up behind quite trivial handling processes. The answer is to make getting defective items back for repair as high a priority task as that of getting serviceable items to the locations at which they can be used.

To make that task easier, computerised asset-tracking is sometimes employed.

_________

Published: Purchasing and Supply Management, February 1989, "Repairables: strategy & management"