Industrial IoT Factory Anomaly Detection Dataset

This dataset focuses on operational integrity within IoT-driven factories. It includes 15,000 instances recorded from various sensors monitoring ambient temperature, light intensity, motion detection, and power consumption. The data is specifically designed to train models in identifying anomalies such as equipment malfunctions, power surges, and security breaches (unauthorized motion). It features highly imbalanced classes (17.4% anomaly prevalence), making it an excellent benchmark for testing robust machine learning algorithms like Logistic Boosting, Random Forest, and SVM. Preprocessing details like SMOTE application and Min-Max scaling are provided to ensure reproducibility in anomaly detection performance metrics.