The Platform
Fault prediction built on electromagnetic signal analysis
Magnefy reads the full EM output of your transformers, compares every reading against a learned baseline and 40+ fault signature patterns, and delivers structured alerts weeks before a failure event.
Core Capability
Continuous EM monitoring. No new infrastructure.
The Magnefy sensor node clips onto existing wiring at the transformer. It captures the electromagnetic signature at 10-second intervals and transmits over your existing cellular or network connection to our processing pipeline. No separate power supply. No trench work. Installation takes one field hour per unit.
The sensor samples voltage and current waveforms across all three phases, capturing harmonic content up to the 25th order. That full spectrum is what lets us see winding insulation changes that thermal sensors miss entirely for weeks.
Signature Analysis
40+ fault patterns, per-fault confidence scoring
Each fault type leaves a distinct harmonic fingerprint. Winding insulation degradation changes the third and fifth harmonic ratios. Core saturation shifts the fundamental frequency envelope. Cooling system faults alter the load-temperature relationship across the EM profile.
Our models were trained on field data from oil-filled distribution transformers across multiple voltage classes. Each alert includes the fault type, confidence score, severity level (watch, caution, action), and the predicted event window so maintenance teams can schedule accordingly.
- Winding insulation degradation Harmonic distortion in phases 1-3, 3-6 week lead time
- Inter-winding short circuit Asymmetric flux pattern, 2-4 week lead time
- Core anomaly Magnetization current deviation, 4-8 week lead time
- Cooling system fault Thermal-load curve departure, 1-3 week lead time
Alert Workflow
Structured alerts that go into your existing workflow
When the model detects a developing fault pattern, we generate a structured alert and push it to whatever endpoint your team uses: SCADA via OPC-UA, CMMS via REST webhook, or plain email digest. The alert carries everything a maintenance coordinator needs to schedule the job.
No new dashboard to check. No separate login for your operations team. Magnefy inserts into the workflow that already exists rather than adding a parallel one.
Integrations
Connects to what your team already runs
Alerts push into your existing systems. No custom middleware, no manual data export.
SCADA / OPC-UA
Native OPC-UA server publishes fault alerts as structured nodes your SCADA can subscribe to directly
REST Webhooks
JSON payloads push to any CMMS endpoint (Maximo, SAP PM, Infor) when a fault alert fires or severity changes
Email Digest
Daily or immediate email alerts formatted for maintenance coordinators, configurable per asset group
CSV / FTP Export
Scheduled flat-file exports for teams with batch-import CMMS or internal reporting pipelines
REST API
Pull asset health scores, active alerts, and historical trend data into internal dashboards or BI tools
NERC CIP Compatible
Data handling architecture reviewed against NERC CIP-007 and CIP-011 requirements for utility environments
Pilot Results
Numbers from the 20-transformer municipal pilot
3-month evaluation at a Pennsylvania municipal electric utility, Q4 2025. Results independently documented in the pilot closure report.
94%
Fault detection accuracy across all 20 monitored assets
17 days
Average advance warning before the predicted failure event
0
False positives that triggered an unneeded maintenance dispatch
Source: Pilot results report, Pennsylvania municipal electric utility, Q4 2025. Available on request for qualified prospects.
See Magnefy on your transformer fleet
A 30-minute technical walkthrough with our engineering team. We review your fleet profile and show you what the platform surfaces on similar assets.