Open Source • MIT License

Predictive Maintenance
for the Modern Factory.

Stop downtime before it happens. The open-source standard for IIoT anomaly detection.

Enter ConsoleView on GitHub
42
https://smart-energy-guardien.io/dashboard
Uptime
99.9%
Devices
847
Alerts
3

Built for Scale.
Designed for Speed.

Enterprise-grade features that just work. No configuration hell.

AI-Powered Detection

Isolation Forest Algorithm

Automatically detect anomalies in real-time sensor data with machine learning. No manual thresholding required.

VIBRATION ANALYSIS
ANOMALY
99.2%
Accuracy
<50ms
Latency
10k+
Predictions/sec

Real-Time MQTT

Stream data from thousands of devices with sub-second latency.

LIVE847 devices
sensors/temp/1 → 72.1°C
sensors/temp/2 → 72.2°C
sensors/temp/3 → 72.3°C

Secure & Scalable

Enterprise-grade security with JWT auth and role-based access control.

End-to-end encryption
Auto-scaling
99.99% SLA
Developer First

Built with
Modern Python.

FastAPI backend with async Python. Powered by scikit-learn for ML inference. Deploy anywhere with Docker.

FastAPI
Async REST API with OpenAPI docs
scikit-learn
Production-ready ML models
PostgreSQL
Time-series optimized storage
Redis
Real-time caching layer
View Documentation
anomaly_detector.py
1from sklearn.ensemble import IsolationForest
2import numpy as np
3 
4class AnomalyDetector:
5    def __init__(self, contamination=0.1):
6        self.model = IsolationForest(
7            contamination=contamination,
8            random_state=42,
9            n_estimators=100
10        )
11    
12    def train(self, X: np.ndarray):
13        """Train the anomaly detection model"""
14        self.model.fit(X)
15        return self
16    
17    def predict(self, X: np.ndarray):
18        """Predict anomalies (-1) or normal (1)"""
19        predictions = self.model.predict(X)
20        scores = self.model.score_samples(X)
21        
22        return {
23            'anomalies': predictions == -1,
24            'scores': scores,
25            'confidence': np.abs(scores)
26        }
27 
28# Real-time inference
29detector = AnomalyDetector(contamination=0.05)
30result = detector.predict(sensor_data)
31
Python 3.10
UTF-8
Ln 24, Col 8

Powered by Industry-Leading Technologies

Next.js
FastAPI
Docker
PostgreSQL
InfluxDB
MQTT
Tailwind CSS
TypeScript
Python
scikit-learn
Redis
Grafana
Next.js
FastAPI
Docker
PostgreSQL
InfluxDB
MQTT
Tailwind CSS
TypeScript
Python
scikit-learn
Redis
Grafana
1.2k+
GitHub Stars
847
Production Deployments
3.5k+
Community Members
42
Countries

Ready to prevent downtime?

Join hundreds of factories already using Smart Energy Guardien.

Let's Get StartedView Documentation