Intelligent predictive maintenance: a bibliometric approach through the SCOPUS database

Intelligent predictive maintenance is an important technique to increase the efficiency and safety of the industry, since it allows detecting and preventing machine problems before they occur. Objective: This study aims to evaluate the scientific production and its evolution over time by means of a bibliometric analysis. Methodology: The search was carried out in the Scopus and WoS databases. The R package Bibliometrix was used to determine the production, impact and collaboration indicators. Statistical software such as SPSS and UCINET were also used to analyze the main approaches. Results: 24 publications were found between 2011 and 2022, with authors Li. Z and Chiu. Y-C being the most relevant in the field. Topics identified as relevant but underdeveloped include "Deep Learning", "Artificial Intelligence", "Big Data Analytics", "Predictive Maintenance", "Industry 4.0" and "Intelligent Predictive Maintenance". Conclusions: As future perspectives in the research, the incorporation of additional techniques such as Bayesian networks, hidden Markov models, and Monte Carlo simulation have been identified. Also, the integration of historical machine operation and failure and maintenance data, along with condition monitoring data, into the data analysis has been proposed. Value: The findings of the study were presented with the intention of being useful to the scientific community.