Smart infrastructure systems combine power-grid sensors, traffic counters, building controllers, weather feeds, maintenance logs, and public-service data into continuously changing operational streams. These systems require more than high-throughput ingestion: they need low-latency anomaly detection, reliable forecasting, cross-source evidence linking, and data-quality controls that respect operational technology constraints. This paper presents Infrastructure Data Intelligence Loop (IDIL), an edge-cloud processing architecture for smart infrastructure analytics. IDIL represents assets as a typed sensor graph, performs early quality checks and compact summaries at edge sites, applies stream-time alignment and concept-drift monitoring in the regional cloud, and gates model actions by confidence, agreement, and evidence completeness. The framework is designed for operational uses such as phasor measurement monitoring, traffic-flow prediction, building-energy anomaly detection, and maintenance-event triage. A prototype evaluation over three representative workloads shows that IDIL reduces median event-to-decision latency by 41.8%, lowers wide-area bandwidth by 36.5%, improves anomaly F1 from 0.71 to 0.79, and reduces stale forecast alarms by 28.6% compared with centralized stream processing. The results indicate that intelligent smart-infrastructure processing is most effective when edge summarization, temporal modeling, evidence retrieval, and governance checks are designed as one loop.
Kumar et al. (Sun,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: