Decision Support System for Automatic Preparation of Integrated Predictive Maintenance Plans

Offentlig prosjektsammendrag

During their lifetime, process plants (example, offshore rigs, chemical plants) or infrastructures (example, bridges, electrical powerlines, railway tracks) are subjected to a number of environmental and operational attacks resulting in their failures. These failures not only cause losses to operators in terms of reduced production and rehabilitation, but more importantly pose significant health, safety and environment (HSE) hazards. Hence, these structures need to be protected, inspected and maintained (repair or replacement) to prevent their failure. Today, most of the asset integrity management strategies are based on experience, risk analysis (example, risk-based inspection (RBI)) or reliability analysis (example, reliability-centered maintenance (RCM)). These approaches provide methods to judiciously divide resources so as to allocate more resources to maintaining critical assets. Recent advances in Information and Communication Technology (ICT) allows data to be acquired, stored, organised, transmitted and converted into useful information at a level that has been unimaginable even in the recent past. Unfortunately, the stages involving data interpretation have not kept pace with data collection, storage and transmission. As a result, there is no generic system that can take a holistic approach to using the available data and managing individual steps to support decision-making. Hence, there is a need for a system that can provide a decision support whereby the data is collected, transmitted and analysed so that it can be used for maintaining the equipment or structure. Thus, the strength of the computer (computation speed, accuracy, almost unlimited storage capacity and analytics) is used to replicate the strength of humans (decision making) to come up with the best solution(s). The project aims to integrate RCM, RBI and condition monitoring approaches into one single platform to develop optimised Predictive Maintenance programs.