Edge AI · Industrial Motor Protection

AI that stops a motor before it fails

This demo simulates an NVIDIA Jetson Edge AI node monitoring a VFD-controlled motor in real time. Adjust load, ambient temperature, and current — watch the AI predict overcurrent and trip the motor before damage occurs. No cloud round-trip. Sub-second inference.

Edge Node: Online

Protocol: Modbus RTU

AI Inference: Active

Motor Status ● RUNNING

480V / 3Ø

→

VFD

→

1750

RPM

NVIDIA Jetson
Edge AI Node

⇅

AWS Dashboard
+ Telemetry

Motor Current (A) 12.4 A

0 AFLA: 28A40 A

Winding Temperature (°C) 62°C

20°CLimit: 120°C150°C

VFD Output Frequency (Hz) 58.3 Hz

0 HzRated: 60 Hz65 Hz

Power Factor 0.87

Simulation Controls MODBUS REGISTERS

🤖

AI Overcurrent Prediction

No risk detected

Confidence: 94% · Last inference: 0.08s ago

Mechanical Load (%) 45%

Simulates torque demand from the driven load (pump, conveyor, compressor). Higher load draws more current from the motor.

Ambient Temperature (°C) 35°C

High ambient temperature reduces motor cooling efficiency, accelerating winding temperature rise under load.

Supply Voltage Imbalance (%) 1.2%

Phase voltage imbalance causes disproportionate current increase in one phase — a leading indicator of overheating.

Real-Time Current History + AI Prediction Window 10-second window · 100ms sampling

Actual Current AI Predicted Trip Threshold

Protection Settings Editable

Full Load Amps (FLA)

Amperes

Trip Threshold

% of FLA

AI Pre-Trip Warning

% of FLA

Prediction Window

Seconds ahead

AI Event Log ● Live

00:00:00Edge AI node initialized. Motor parameters loaded.

00:00:01Modbus RTU connection established. Polling at 100ms.

00:00:02Motor started. Baseline current: 12.4A. Status: Normal.