Modeling Sensor Networks for Predictive Maintenance

Predictive Maintenance is one of the most intensively investigated topics in the current Industry 4.0 movement. It aims at scheduling maintenance actions based on industrial production plants’ past and current condition and therefore incorporates other trending technological developments such as the Internet of Things, Cyber-Physical Systems or Big Data Analytics. In this short paper we motivate the employment of machine learning algorithms to detect changing behavior as indication for the necessity of maintenance on a microscopic level and describe how cyber-physical environments benefit from this approach.