Streamed Meter Data Violation Detection System

The patent segments the batch processing approach into two parallel paths: a streaming processing path that continuously analyzes incoming meter data in real-time to detect violations immediately, and a batch processing path that periodically processes accumulated data to ensure completeness. This segmentation allows the system to achieve both fast violation detection through streaming and comprehensive data processing through batch operations, resolving the contradiction between response time and processing completeness.

The patent introduces an intermediary component called a 'stream processor' that sits between the data reception layer and the violation detection logic. This stream processor maintains an in-memory cache of recent meter readings and implements a sliding window mechanism to efficiently compare current readings against historical data. By using this intermediary with optimized data structures and algorithms, the system achieves real-time violation detection without requiring complex distributed computing frameworks, thus balancing detection speed with manageable system complexity.

The patent extracts and processes only the most relevant features from meter data in real-time, such as current consumption values, rate of change, and comparison with predefined thresholds or customer-specific profiles. Instead of processing and storing all raw meter data, the system extracts key indicators of potential violations and processes only these extracted features. This extraction approach maintains high detection accuracy by focusing on critical data points while significantly reducing the overall data processing load and memory requirements.