Applying Data Science to Improve Asset Management
June 4 and 5, 2024, in Huelva, Spain. Organized by AEM – Spanish Maintenance Association. Phoenix AMC’s Paper chosen by a Selection Committee.
Asset management is increasingly relevant in optimizing operational performance, reducing costs and risks, and extending the useful life of assets. The use of statistical tools enabled by data science is presented as a great opportunity to improve efficiency and effectiveness in decision making, thus optimizing asset management.
In this article, Phoenix AMC wishes to share its knowledge and advances in the integration of statistical tools that allow the optimized treatment of multiple variables, through the use of data intelligence in asset management.
The methodology is based on advanced analytical tools for the automation of tasks, monitoring and optimization of equipment performance, including the optimization of energy use and therefore assisting decarbonization. Some of the tools we use are: Digitization of information from assets, Weibull analysis to identify potential opportunities for improvement, Mahalanobis analysis to detect anomalies (unusual patterns and atypical behaviors) and others.
This methodology has wide applications. In the specific case of maintenance in the chemical and process industry, it is used to predict potential failures before they occur, detecting early signs of equipment degradation. This enables the ability to proactively schedule maintenance activities, increasing availability, performance and equipment life.
The implementation of this methodology was achieved through computational modeling supported by the Python programming language, which facilitates the automation of analysis and decision-making tasks in asset management.
