The Grid4C engine monitors each meter separately, automatically learns its underlying correlations, and uses unique information-theory based algorithms in order to decompose each meter's behavior into sub-series, which are then automatically modeled. The exclusive ‘problem decomposition’ feature visually presents the meters 'personalized behavior rules' in a way that can be easily understood by both energy consumers and energy providers. These rules enable to extract real-time actionable insights, perform various types of 'what if' analysis, in a manner that is intuitive and intelligible to the end user. After the modeling phase, the Grid4C engine monitors each meter,
using distinctive anomaly detection algorithms, in order to detect early warnings of changes in the meters' patterns and behavior. Whenever such a deviation is detected, the Grid4C solution automatically identifies the irregularities' signature, and determines whether irregularities are caused due to bypass and tampering, meter malfunctioning, or change in consumption patterns. Whenever changes in consumption behavior are identified, changes are classified and real-time actionable insights are extracted. The result is a full self-learning adaptive mechanism, guaranteeing accurate predictions, detection of anomalies at an early stage, and real-time actionable insights.