Multi-Tenant SaaS With Role-Optimized Workflow
Fleet Management IoT Platform
Multi-tenant
with pluggable integrations
Role-optimized
dispatcher / manager / driver views
Industrial IoT / Medium-Voltage Infrastructure
13kV RMU Fault Detection System
Partial-Discharge Sensing with Evidence-Rich Alerts
Industrial IoT
Industry
9 months
Duration
2 → 8
TRL
8
Disciplines
Background
A utility operator faced high costs and customer impact from ring main unit faults detected only after outages occurred. They required a retrofit system that could identify incipient faults before failure.
The problem
What made it hard.
Medium-voltage environments are electrically noisy and physically constrained. The system had to sense meaningfully - partial discharge, thermal anomalies, load imbalance - without interfering with the RMU itself, and had to justify each alert to dispatch engineers.
Approach
What we built.
We engineered a sensor pod with partial-discharge, temperature, and current/voltage measurement, feeding an edge analytics unit that classified signatures against known fault patterns. Alerts carried supporting evidence - trend, signature type, and confidence - to make dispatch decisions defensible.
Workstreams
What the program actually covered.
This was delivered as 4 coordinated workstreams over 9 months.
WS 01
Real-Time Electrical Fault Analytics Platform
We built a streaming analytics pipeline on a purpose-built time-series store, with anomaly detectors running per feeder and aggregating to substation-level views. Alert thresholds are learned per asset rather than fixed.
WS 02
High-Voltage Fault Classification System
We combined DSP feature extraction with ML classification, trained against a curated waveform dataset. Models were calibrated per substation where needed, with a fallback to generalized models for new sites.
WS 03
Smart Substation Monitoring System
We designed an edge gateway that sits alongside existing protection, consuming data via IEC 61850 and legacy protocols without disturbing them. Locally, the gateway serves SCADA needs; remotely, it pushes telemetry to cloud analytics.
WS 04
Electrical Grid Monitoring Dashboard
We designed a GIS-first layout with drill-downs into feeders, assets, and events. A unified timeline connects live and historical views, and customizable layers let different operator roles tailor what they see.
Outcome
What it measured.
Early
fault detection before outage
Evidence-rich
alert payload
Retrofit
to existing RMUs
The system flagged incipient faults in pilot deployments that would otherwise have progressed to outages. Field engineers reported that the evidence attached to each alert made triage fast and reduced false-positive fatigue.
Our role
Sensor design; edge analytics; field commissioning.
Technologies
Partial-discharge sensingcurrent/voltage meteringedge AITime-series DBstreaming analyticsanomaly detectionDSPML classifierswaveform databasesIEC 61850 awarenessedge gatewaycloud telemetry
Gallery
Inside the build.
Illustrative of the sensor node's front end: the partial-discharge input stage and the current/voltage metering channels sit on a board of this class inside the ring main unit enclosure.
Representative of the retrofit constraint: sensing had to be added inside existing, physically cramped medium-voltage assemblies during scheduled outage windows.
Illustrative of the bench work behind the discharge classifier - separating genuine PD pulses from switching and corona noise started with captured waveforms long before any model was trained.
Next case study
Tell us what you’re building.
Send the constraint that worries you most - a latency budget, a power budget, a certification date. We’ll tell you straight whether we’re the right team.