GridSense
The grid, seen clearly, half an hour at a time.
- Client
- Voltane Energy
- Year
- 2025
- Duration
- 12 months
- Team size
- 9 engineers
Interactive Demo
A hands-on mock of the product we built — click around.
Demo interface shown in English.
The challenge
Voltane’s generation fleet had tripled in three years, but each site still reported through its own SCADA historian, its own units and its own idea of time. Traders priced positions on half-hourly settlement data that finance reconciled manually — a fortnight in arrears, with disputes worth six figures a quarter. Faults surfaced as phone calls from maintenance crews, not alerts; one transformer ran hot for eleven days before anyone noticed, and the repair cost forty times what early intervention would have. Ofgem reporting obligations were tightening, and the board wanted fleet-level visibility before committing to another £200m of capacity. Data existed everywhere and informed nothing.
Our solution
We built an ingestion layer that speaks MQTT and OPC-UA at the edge, normalises every feed into a shared schema and lands 2.4 million readings an hour in TimescaleDB. A WebSocket-driven React dashboard gives operators a live map of the fleet, with drill-down from portfolio to a single inverter in three clicks. Python anomaly-detection models — one per asset class, retrained weekly — flag drift, underperformance and thermal signatures hours to days before failure, routed to the right engineer with context attached. Settlement reconciliation now runs automatically against metered data, producing audit-ready Ofgem reports, and the whole stack runs on Azure Kubernetes Service with infrastructure as code.
Results
Technology stack
Facing a similar challenge?
We've solved this class of problem before. Let's talk about how that experience applies to your situation.
Discuss your projectNext case study
Clausecraft →