Predictive Edge Analytics for Sanitary Facility Monitoring
Project Overview
Sensors deployed in remote sanitary facilities traditionally stream all data to the cloud, incurring significant bandwidth costs, latency, and outage risks, particularly in regions with unreliable mobile network connectivity. This project designs a context-aware distributed analytics architecture that dynamically orchestrates computational workloads across on-device, edge, fog, and cloud resources so that only summarised insights travel upstream to central servers. The research addresses the challenge of maintaining reliable hygiene service monitoring in vulnerable communities where continuous cloud connectivity cannot be guaranteed. The project reviews existing anomaly-detection approaches applicable to sanitary facility sensor data and develops self-organising orchestration algorithms that adapt processing placement based on network conditions and resource availability. By enabling predictive analytics at the edge, the system can detect maintenance needs and usage anomalies locally, triggering appropriate responses without requiring constant upstream data transmission.
The flexible analytics pipeline improves service quality while reducing dependency on mobile networks in remote or underserved areas. The orchestration algorithms dynamically redistribute computational workloads as connectivity conditions change, ensuring that critical monitoring functions continue to operate even during network outages or periods of degraded mobile service.
The work is conducted in collaboration with iPoint, which provides transport and hygiene facility management solutions. Funded through the KESS 2 East programme that pairs Welsh industry organisations with university researchers, the project contributes to understanding how distributed IoT analytics can serve infrastructure monitoring in resource-constrained deployment environments.
Team
Funding
