Cost-effective designs of field service for electronic systems
Field service is a complex process that involves sending technicians to perform repair tasks on consumer products. It is responsible for having failed products fixed, achieving high customer satisfactions, and keeping a low service cost. The costs of field service significantly increase due to the rule of ten. Also, the cost of field service is an important corporate activity but it is often not well understood. Field repair service is the major part of field service and it is well known that the field repair cost is a vast amount of money. It is a serial of diagnostics to decide whether failed products are to be repaired or discarded. Minimal the field repair cost is particularly critical to global competitiveness of an electronic company. Trading off between the significance of impact on performance and manageability of problem complexity as a first step, we focus on two systemdesign level problems, alternative maintenance policies and field repair service, in this dissertation. Exploiting the knowledge-based technology as well as the multiobjective analysis, we develop a decision tree selecting model (DTSM) for alternative maintenance policies, and further, design a weight-based maintenance policy (WBMP) for personal computer (PC) systems. Cores to the knowledgebased policy selecting are decision tree representation models. This design identifies maintainability problems early in the system design and selects economic maintenance policies for electronic systems. The key of the WBMP system is to use product's status as a guide for performing maintenance tasks. A simulation model based on a field service operation is built to estimate the value of the WBMP system. Test runs of DTSM and WBMP over field service compatible environments demonstrate both the feasibility and potential effectiveness for applications. For field repair test, we present a decision tree diagnosis tool to reduce the cost of test. We also design a cost-effective approach for overall system-level test, which incorporates a manufacturing system test model, the field failure rate analysis, and a field repair model into a test cost model. The optimized system design parameters from the cost model can be used to reduce the cost of systemlevel test.