A Review of IoT-Enabled Predictive Maintenance in Industrial Automation: Opportunities, Challenges, and Future Directions

The field of predictive maintenance enabled by the concept of Internet of Things has become an increasingly important research area, especially considering the evolution of industrial systems from traditional architectures to connected and data-driven modes of operation. Thus, distributed nodes collect sensory data, including vibration, temperature, electrical, and environmental data, whereas data-driven analytical approaches facilitate early indications of possible failures and estimation of useful life. This paper presents a survey of recent research trends related to the development of predictive maintenance frameworks enabled by the Internet of Things across different application domains, including manufacturing, energy systems, mining, and smart-building environments. Based on this survey, several promising research directions and key benefits of predictive maintenance enabled by the Internet of Things are identified, including real-time operational monitoring, early detection of anomalous behavior, and more efficient planning and scheduling of maintenance activities. Nevertheless, despite these promising advantages, there are some limitations that can be identified, for instance, the ones relating to the data quality, potential communication latency, security, and compatibility with conventional industrial systems. This survey also identifies potential research directions to enhance the reliability, connectivity, security of the systems, for the effective implementation of the predictive maintenance systems.

کلیدواژه ها:

Deep Learning ، Industrial Internet of Things ، Internet of Things ، Machine Learning ، Predictive Maintenance