Knowledge-based Estimation of Maintenance Time for Complex Engineered-to-Order Products Based on KPIs Monitoring: A PSS Approach

Abstract

Maintenance is the most common product-oriented Product Service System (PSS) offering, as well as a core activity of the manufacturing system lifecycle, since it accounts for as much as 60 to 70% of its total costs. Commonly, in the Small Medium Enterprises (SMEs), the maintenance activities of customised engineering-to-order (ETO) products highly depend on the experience of engineers and shop-floor experts, without considering specific tools and algorithms that can capture the knowledge and reuse it in an efficient way. Moreover, the estimation of the maintenance time for a new maintenance project, which is among the main offerings in the maintenance, is solely based on the engineer's experience and knowledge. Aiming to support the knowledge capturing and its reuse in the maintenance activities, as well as to improve the performance of the provided maintenance PSS, the present work proposes a methodology for knowledge-based estimation of maintenance time based on Key Performance Indicators (KPI) monitoring. Data captured through the KPIs monitoring tool were collected in a knowledge repository, and were processed using a Case-Based Reasoning (CBR) technique, estimating the required maintenance time. A validation of the proposed methodology was performed based on real-life data from a mold-making European Small Medium Enterprise (SME). Preliminary results indicated a significant reduction in the number of iterations between customers and the engineering department, compared to the traditional approach followed by the company, and improved accuracy of maintenance time estimation, which led to increased customer satisfaction.