Imbalance Detected
Severity: 0.92
RUL Estimate
44 days remaining
API Status
Online ✓
Last sync: just now
Vibro Platform — Live Monitoring Active

AI Vibration Analysis Platform

Real-time vibration monitoring across 101 machine types in 7 industrial categories. 15 ML models detect 12 fault categories in parallel with sub-7ms inference. From sensor to insight in milliseconds.

Get Started View Platform Demo
101
Machine Types
12
Fault Categories
15
ML Models
<7ms
Inference
Scroll
Fault Detection

12 Fault Categories Detected in Real-Time

Covering 101 machine types across 7 industrial categories — from motors and pumps to turbines and crushers. Each fault mapped to specific machines with priority levels.

Machine Learning

15 ML Models Powering Detection

12 per-fault production models backed by 3 training architectures — CNN-LSTM, 1D CNN, and VMD-CNN — all running simultaneously for comprehensive coverage.

Anomaly Detection

CNN-LSTM Hybrid

Dual-branch architecture processing 24 features through FFT and time-domain branches. Fusion layer classifies all 12 fault categories with 99.9% accuracy.

12 Classes PyTorch ONNX
Fault Classification

1D CNN Classifier

Raw waveform input — no manual feature extraction needed. Processes 2048-point signals directly through 3 convolutional layers for instant fault identification.

Raw Waveform 99.9% Acc Real-Time
RUL Prediction

VMD-CNN + LSTM

Variational Mode Decomposition separates signal modes before CNN processing. Combined with LSTM for Remaining Useful Life prediction in days.

Signal Decomp RUL (Days) SHAP XAI
Platform

Live Vibration Analysis Dashboard

From sensor to insight in milliseconds. Real-time monitoring with AI-powered fault detection.

🔬
101 machine types covered
7 industrial categories — Rotating Machines, Process Machinery, Drive Train, Heavy Industrial, Power Generation, HVAC, and more. Each mapped to applicable fault types with priority levels.
🧠
15 ML models running in parallel
12 per-fault production models plus 3 shared classifiers (CNN-LSTM, 1D CNN, VMD-CNN). Autoencoder for anomalies, classifier for faults, LSTM for remaining useful life prediction.
⚡
<7ms real-time inference
FastAPI backend with continuous 24/7 processing. 25,600 Hz sampling rate ensures no fault goes undetected.
🔔
Instant multi-channel alerts
Email and WhatsApp notifications within seconds of fault detection, including fault type, severity, and recommended action.
Request Demo
vibro.ae/dashboard
Fault Type
IMBALANCE
Anomaly Score
0.9234
RUL
44.3 days
Confidence
96.2%
Normal12%
BPFO28%
Imbalance60%
WARNING — Imbalance detected on Motor #2
API ML DB 25,600 Hz · live
0+
Machine Types
0
ML Models
<0ms
Inference Time
0%
Platform Uptime
Pipeline

From sensor to insight
in milliseconds

A fully automated vibration analysis pipeline — no human intervention required until an alert is issued.

01
Data Capture
Vibration Sensor
Accelerometer captures vibration at 25,600 Hz. Continuous high-resolution signal from bearings, motors, and rotating equipment.
02
Feature Extraction
Signal Processing
15+ statistical and frequency-domain features extracted in real time: RMS, Kurtosis, Crest Factor, Skewness, and envelope spectrum.
03
AI Inference
3 ML Models
15 ML models work together — CNN-LSTM classifies faults, 1D CNN processes raw waveforms, VMD-CNN decomposes signals, and LSTM predicts Remaining Useful Life.
04
Instant Notification
Alert
Email and WhatsApp alerts within seconds. Fault type, severity score, affected asset, and recommended action included automatically.
Get Started

Ready to monitor
your vibrations?

Tell us about your equipment and monitoring needs. We respond within one business day.

Message Sent!

Thanks for reaching out. We'll be in touch within one business day.

info@vibro.ae vibro.ae