Data science and machine learning in the IIoT concepts of power plants

Smart power plants are no longer just futuristic ideas but are rapidly becoming a reality, and for this to be achievable, an application of artificial intelligence in Industrial Internet of Things (IIoT) concepts is necessary. This paper presents the place and role of Data Science (DS) and Machine Learning (ML) on edge, fog, and cloud levels of vertical IIoT concepts of power plants. A comprehensive functional analysis of edge, fog, and cloud levels has been done. Data analyzing, preprocessing, and processing are described on all levels separately. Limitations in signal conversion and data preprocessing at the edge level (edge computing), data processing and analysis at the fog level (fog computing), and information processing at the application and cloud levels (highend and cloud computing) are described in detail. ML algorithms have been selected based on the particular management and control level. The proposed concepts represent management and maintenance improvements with minimal investment and avoidance of production downtime.