Cyber-enabled Product Lifecycle Management: A Multi-agent Framework

Trouble free use of a product and its associated services for a specified minimum period of time is a major factor to win the customer's trust in the product. Rapid and easy serviceability to maintain its functionalities plays a key role in achieving this goal. However, the sustainability of such a model cannot be promised unless the current health status of the product is monitored and condition-based maintenance is exercised. Internet of Things (IoT), an important connectivity paradigm of recent times, which connects physical objects to the internet for real-time information exchange and execution of physical actions via wired/wireless protocols. While the literature is full of various feasibility and viability studies focusing on architecture, design, and model development aspects, there is limited work addressing an IoT-based health monitoring of systems having high collateral damage. This motivated the research to develop a multi-agent framework for monitoring the performance and predicting impending failure to prevent unscheduled maintenance and downtime over internet, referred to as for cyber-enabled product lifecycle management (C-PLM). The framework incorporates a number of autonomous agents, such as hard agent, soft agent, and wave agent, to establish network connectivity to collect and exchange real-time health information for prognostics and health management (PHM). The proposed framework will help manufacturers not only to resolve the warranty failure issues more efficiently and economically but also improve their corporate image. The framework further leads to efficient handling of warranty failure issues and reduces the chances of future failure, i.e., offering durable products. From the sustainability point of view, this framework also addresses the reusability of the parts that still have a significant value using the prognostics and health data. Finally, multi-agent implementation of the proposed approach using a power substations for IoT-based C-PLM is included to show is efficacy. (C) 2019 The Authors. Published by Elsevier Ltd.