IIoT Edge Computing Dataset for Predictive Maintenance and Real-Time Control

Dataset Overview

This dataset is specifically designed for research on real-time edge computing architectures in Industrial Internet of Things (IIoT) applications. It simulates sensor data, network latency, and predictive maintenance scenarios in a smart manufacturing environment, integrating advanced Fuzzy PID controllers for adaptive industrial process control.

Key Features

• Time-series IIoT sensor data including temperature, pressure, and vibration measurements
• Edge computing performance metrics: network latency and processing time
• Fuzzy PID controller output optimized for real-time control applications
• Maintenance status classifications: Normal, Warning, Failure
• Predictive failure labels (binary: 1=Failure, 0=No Failure) for supervised learning
• Engine cycle data for remaining useful life (RUL) estimation
• Test data without RUL labels for model evaluation
• Ground truth RUL values for performance validation

Data Structure

The dataset is organized into multiple components:

• Sensor Measurements: Temperature (°C), Pressure (bar), Vibration amplitude (mm/s)
• Edge Computing Metrics: Network latency (ms), Processing time (ms), Data transmission rates
• Control System Data: Fuzzy PID controller outputs, Setpoint values, Error signals
• Maintenance Information: Current status (Normal/Warning/Failure), Maintenance schedules
• Predictive Labels: Binary failure prediction, Confidence scores
• Temporal Features: Timestamps, Engine cycles, Operating hours
• Ground Truth: Actual RUL values for test engines

Data Collection Method

The dataset simulates a realistic smart manufacturing environment where IoT sensors continuously monitor industrial equipment. Data generation incorporates realistic noise patterns, sensor drift, and failure progressions. The Fuzzy PID controller simulations are based on industrial control theory, providing authentic responses to varying operational conditions and edge computing constraints.

Research Applications

• Real-time edge computing architecture optimization for IIoT
• Predictive maintenance model development and evaluation
• Anomaly detection in industrial sensor networks
• Fuzzy PID controller tuning and performance analysis
• Network latency impact on industrial control systems
• Remaining useful life (RUL) prediction for rotating machinery
• Autonomous decision-making in smart manufacturing
• Edge vs cloud computing tradeoff analysis for IIoT

Machine Learning Use Cases

• Binary classification for failure prediction
• Multi-class classification for maintenance status
• Time series forecasting for sensor values and RUL
• Regression models for RUL estimation
• Deep learning (LSTM, GRU, Transformer) for sequential data
• Anomaly detection using autoencoders and isolation forests
• Reinforcement learning for adaptive PID tuning
• Feature engineering from vibration and sensor fusion
• Real-time streaming analytics and edge inference