Real-Time Streaming Detection for Water & Energy Network Leaks & Faults
SeafaringIT
StageSur site4 à 6 moisDate limite : 23 févr. 2026
Streaming DataBackend microservices (Nest, Kafka)Flink/Spark StreamingAnomaly detectionbackend APIsTime-series AnalysisDevopsIndustrial Dashboards
Description
- Goals:
- Build a real-time streaming pipeline to detect leaks and anomalies in water/energy networks
- Enable near-instantaneous detection to reduce losses and risks
- Provide actionable insights for supervision and maintenance
- Student roles: Data engineers, streaming specialists, ML engineers, backend engineers, DevOps engineers
- Expected outcomes: Real-time streaming platform with sensor ingestion, anomaly detection, contextual correlation, dashboards, and operational procedures
- Key features:
- Real-time ingestion from sensors (simulated and real)
- Streaming anomaly detection with scoring and prioritization
- Contextual correlation (zones, historical patterns, weather)
- Real-time supervision API and dashboards
- Alert management and escalation
- Performance monitoring (latency, false positives)
- Operational runbooks; integration with SCADA/monitoring systems
- Technologies: Apache Kafka, Apache Spark/Flink, anomaly detection algorithms, stream processing, PostgreSQL/TimescaleDB, Redis, FastAPI, BI tools (Grafana, Kibana), Docker, Kubernetes