Real-time transformer health.
HVTM is an AI-powered health monitor for distribution transformers. Clamp-on sensors capture vibration, acoustics, temperature and load, while the cloud platform detects outages, scores anomalies and flags developing faults — before they take the lights out.
HVTM-001
Distribution transformer · 100 kVAIllustrative dashboard preview
Solutions
An end-to-end stack — rugged sensing hardware, resilient connectivity, and AI analytics — built for the realities of African distribution networks.
Multi-Signal Sensing
A clamp-on sensor node captures everything a transformer is telling you — no outage, no drilling.
- Vibration & contact acoustics
- Airborne audio
- PT100 temperature & CT load
- Tamper detection
AI Health Analytics
Machine-learning models built for distribution assets score anomalies and classify developing faults.
- Per-asset learned baselines
- Load- and temperature-aware models
- Fault-family classification
- Transformer-level energy balance
Outage Detection & Alerts
Know about an outage in minutes — not when customers start calling.
- Transformer-level outage detection
- Real-time alerts
- Prioritized field-intervention lists
- Event history per asset
Cloud Platform
One dashboard for your whole fleet, from a single site to a national network.
- Live fleet dashboards
- LTE/NB-IoT connectivity
- Exportable reports
- API access
For utilities, mini-grid operators, and commercial & industrial sites.
How it Works
From clamp-on sensors to prioritized action — four steps to full visibility over your transformer fleet.
Clamp On
Install the HVTM node on a live transformer in under an hour — non-invasive, no outage, no modification to the asset.
Stream Telemetry
The node streams vibration, acoustic, thermal and load data over LTE/NB-IoT to the GridSemi cloud platform.
Learn & Detect
Models learn each transformer's normal behavior, then flag anomalies, developing faults, and outages as they emerge.
Act Early
Your team gets prioritized, actionable alerts — fix small faults on a schedule instead of replacing failed transformers in a crisis.
Transformer Health, Live
Catch developing faults before they become failures. HVTM analytics flag excursions, trends, and emerging fault signatures across every monitored asset — for about 2% of what replacing a failed transformer costs.
Anomaly scoring across six signal types
Vibration, contact acoustics, airborne audio, temperature, load, and tamper — fused into one health score
Per-asset baselines, not generic thresholds
Each transformer learns its own normal, conditioned on load and ambient temperature
Prioritized field-intervention lists
Faults, overloads and suspected losses ranked so crews fix what matters first