A predictive maintenance approach based in big data analysis

With the evolution of information systems, the data flow escalated into new boundaries, allowing enterprises to further develop their approach to important sectors, such as production, logistic, IT and especially maintenance. This last field accompanied industry developments hand in hand in each of...

Full description

Bibliographic Details
Main Author: Silva, João Pedro Gonçalves da (author)
Format: masterThesis
Language:eng
Published: 2020
Subjects:
Online Access:http://hdl.handle.net/10071/20241
Country:Portugal
Oai:oai:repositorio.iscte-iul.pt:10071/20241
Description
Summary:With the evolution of information systems, the data flow escalated into new boundaries, allowing enterprises to further develop their approach to important sectors, such as production, logistic, IT and especially maintenance. This last field accompanied industry developments hand in hand in each of the four iterations. More specifically, the fourth iteration (Industry 4.0) marked the capability to connect machines and further enhance data extraction, which allowed companies to use a new data-driven approach into their specific problems. Nevertheless, with a wider flow of data being generated, understanding data became a priority for maintenance-related decision-making processes. Therefore, the correct elaboration of a roadmap to apply predictive maintenance (PM) is a key step for companies. A roadmap can allow a safe approach, where resources may be placed strategically with a ratio of low risk and high reward. By analysing multiple approaches to PM, a generic model is proposed, which contains an array of guidelines. This combination aims to assist maintenance departments that wish to understand the feasibility of implementing a predictive maintenance solution in their company. To understand the utility of the developed artefact, a practical application was conducted to a production line of HFA, a Portuguese Small and Medium Enterprise.