Active Monitoring using Real-time Metric Linear Temporal Logic Specifications

Monitoring temporal relationships among events in event streams has wide scale applicability in health information systems. From detecting violations of privacy policies in message sequences to diagnosing conditions in physiological data streams real-time event monitoring of temporal invariants is becoming an important tool for system design. We developed an Active Real-Time Event Monitoring and Integration System (ARTEMIS) capable of integrating event streams and monitoring the existence of temporal invariants among events expressed in a safety fragment of metric first-order temporal logic (MFOTL). The paper discusses the mathematical foundations of the monitor, and demonstrates the application concepts in a physiological alarm generator and clinical information workflow system.