A real-time IoT vehicle sensor analytics and anomaly detection system.
VehiclePulse simulates 30 days of vehicle sensor data across 5 parameters (speed, engine temperature, fuel level, RPM, battery voltage) and detects anomalies using Isolation Forest.
- Python, Pandas, NumPy
- Scikit-learn (Isolation Forest)
- Matplotlib, Seaborn
| Metric | Score |
|---|---|
| Overall Accuracy | 98% |
| Anomaly Recall | 77% |
| Anomaly Precision | 76% |
| Total Readings | 8,640 |
- Simulated IoT sensor data with injected fault scenarios
- Detects engine overheating, battery drops and speed spikes
- 7-panel analytics dashboard with full sensor telemetry
- Hypothesis-tested anomaly thresholds per sensor type
generate_data.py— Simulates 30 days of vehicle sensor readingsexplore.py— Sensor overview and trend visualizationanomaly.py— Isolation Forest anomaly detection and evaluationdashboard.py— 7-panel analytics dashboard
