Track: Production Planning and Management

Abstract

For manufacturers, a support service is one of the most important factors to maintain customer satisfaction. In order to ensure that repair services are available for users, manufacturers must hold appropriate service part inventories. However, the technologies used for service parts become obsolete over time and production of the parts terminates. Indeed, a service part production period is usually shorter than the period for which a support service is offered. Thus, manufacturers must procure service parts for the period following the termination of service part production. Currently, service part production periods are becoming shorter and manufacturers must hold large inventories of long-term service parts. Thus, for the purpose of holding appropriate service part inventories, an accurate demand forecasting method for service parts is required. In this paper, we propose a new demand forecasting method that is an improvement of an analogy-based method. The improvement is based on the findings of an investigation of service part demand data. Following this investigation, we added a liner regression term into the original analogy-based method. The proposed method was then evaluated using the data provided by a home electrical appliance manufacturer. The numerical experiment showed that the improved analogy-based method achieved higher forecasting accuracy than the conventional method.