A Networked Society based on the Internet of Things is a significant paradigm shift in the early 21st century. The advanced modern engineered systems, constituent of the networked society, within the areas of Utility, Transport, Telecommunication and Enterprise are becoming increasingly dynamic and complex. These encompass various smart devices components, including both software and hardware such as Cyber-Physical Systems. As the number of these components and interactions increases being networked with each other or the internet, it is becoming challenging to manage and operate efficiently their complex networks. Furthermore, these systems can fail, implying impacts to their availability, maintainability, reliability and ultimately customer and end-user satisfaction. Therefore, there is a tremendous need for effective management and operation for both Telecommunications and Industry & Society complex systems, leveraging analytics from Cyber-Physical Systems collected data. In this paper, we propose a generic predictive analysis framework for decision support using a Bayesian Belief Network that will increase the Utility complex systems cost efficiency during the network operations and maintenance lifecycle. The enabling technologies are based on probabilistic and data mining techniques with pattern detection to extract fault precursors leveraging events from the network, communication quality data and trouble tickets. This predictive resolution approach will proactively reduce maintenance cost and improve overall systems management and operations efficiency, performance, reliability and customer satisfaction.
