Analysis of evaluation systems for product repairability: A case study for washing machines

Increasing the durability of electrical and electronics products is key to reduce our worldwide material consumption and the related environmental footprint. A valuable strategy to achieve this goal is extending the useful lifetime of products by means of repair. Methods to assess and rate the repairability of products have been developed recently. In this study, two different assessment methods, the Assessment Matrix for ease of Repair (AsMeR) developed by KU Leuven and the Repair Scoring System (RSS) developed by JRC, are applied to seven washing machine models. The data required for this assessment are collected through a combination of literature study and fieldwork at a refurbishment centre of large household appliances. The aim of this paper is to investigate (1) the ability of the two assessment methods to capture the diversity of products on the market, (2) coherence of results obtained with the methods, and (3) how methodological choices can affect results. The results suggest that there is no perfect correlation between repairability score and product characteristic such as consumer price: cheaper machines in general scored lower in terms of repairability, but the best score was not obtained by the most expensive model. This can possibly be explained by the fact that high-range models are designed for reliability rather than for (self-) repairability. Overall, results show a good coherence between the applied repairability assessment methods. Both methods rely on the selection of priority parts, which is a challenging task. However, the sensitivity analysis revealed that overall scores are not significantly affected by the number and the weighting factor of priority parts, as long as a sufficient number of priority parts are considered (at least five for washing machines). In addition, the number and weighting factor of some criteria and parameters could be reduced without significantly altering the relative ranking of the investigated WMs. This suggests that the methods could be simplified when tailored to assess specific product groups.