Service demand forecasting through the systemability model: a case study

Companies competing in an increasingly competitive market must ensure the production of goods with excellent performance, able to satisfy their customers and which have low manufacturing and management costs. It is in this context that companies have, in recent years, invested in research and development and have upgraded their reliability and maintenance functions. In many cases, the maintenance engineers have attempted to predict the reliability of the products, at least for evaluating the number of warranty repairs to be performed. This approach is on the one hand, extremely appropriate but, on the other, must face the difficulties of making laboratory test in conditions often radically different from those that the products meet during their normal operation. Frequently, the reliability estimation, coming from experimental test (in-house) are different from those obtained by the analysis of the service data (in-field). The former are executed in laboratory with standardized, controlled and repeatable conditions, while the latter are affected by random environmental and operating conditions. In the field of household appliances, this is so true that the conditions of use may vary even from country to country. There are some approaches that allow to assess the reliability performance of a system starting from the results of experimental tests performed in a laboratory. One of these was proposed some years ago and is called systemability. In this study, it was applied, for the first time, this approach to the field of household appliances. In addition, we wanted to try to identify the parameters that allowed to distinguish two different European markets. In fact the in-field data come from two different countries and could be considered a great opportunity to validate the correlation model. In fact, it was possible to investigate the effects of two different environmental condition sets (costumer behaviours, , market issues, logistics, etc.) on the reliability performances of a product population that has been manufactured in the same industrial plant. One of the most important outcomes of the Systemability model was the capacity to predicts two different in-field reliability performances relative to two different markets in contrast with the classic methodology that uses the same in-house reliability data without considering environmental effects. The initial stage of modeling was followed by a second validation phase, which gave satisfactory results. The overall outcomes were very positive and they have allowed us to focus some improvements in maintenance management that will lead to greater effectiveness of the method in the coming years. KeywordReliability, random environmental factor, Systemability, Availability, Maintainability engineering.