CNC-Predictive-Maintenance

This repository contains the code and data for a predictive maintenance system for CNC (Computer Numerical Control) machines. The system aims to predict potential failures and maintenance needs to minimize downtime and enhance operational efficiency.

Table of Contents

• Introduction
• Installation
• Data
• Usage
• Configuration
• Contributing
• License
• Contact

Introduction

Predictive maintenance involves using data analysis tools and techniques to detect anomalies in operations and possible defects in equipment and processes so you can fix them before they result in failure. This project focuses on predictive maintenance for CNC machines, leveraging machine learning algorithms to predict failures based on historical data.

Installation

To run this project locally, follow these steps:

1. Clone the repository:

   git clone https://github.com/Shengwei0516/CNC-Predictive-Maintenance.git
   cd CNC-Predictive-Maintenance

2. Install the required dependencies:

   pip install -r requirements.txt

Data

The dataset used in this project consists of historical data from CNC machines, including sensor readings, operational parameters, and maintenance logs. Below is a description of the main columns in the dataset:

Click to expand

Column	Description
time	Timestamp indicating when the data was recorded.
V_avg_machine	The average voltage supplied to the CNC machine.
I_avg_machine	The average current drawn by the CNC machine.
kW_machine	Active power consumption of the CNC machine in kilowatts.
kvar_machine	Reactive power consumption of the CNC machine in kilovars.
kVA_machine	Apparent power consumption of the CNC machine in kilovolt-amperes.
PF_machine	Power factor of the CNC machine, representing the efficiency of power usage.
kWh_machine	Total energy consumption of the CNC machine in kilowatt-hours.
kvarh_machine	Total reactive energy consumption of the CNC machine in kilovar-hours.
kVAh_machine	Total apparent energy consumption of the CNC machine in kilovolt-ampere-hours.
V_avg_spindle	The average voltage supplied to the spindle.
I_avg_spindle	The average current drawn by the spindle.
kW_spindle	Active power consumption of the spindle in kilowatts.
kvar_spindle	Reactive power consumption of the spindle in kilovars.
kVA_spindle	Apparent power consumption of the spindle in kilovolt-amperes.
PF_spindle	Power factor of the spindle, indicating the efficiency of power usage.
kWh_spindle	Total energy consumption of the spindle in kilowatt-hours.
kvarh_spindle	Total reactive energy consumption of the spindle in kilovar-hours.
kVAh_spindle	Total apparent energy consumption of the spindle in kilovolt-ampere-hours.
RPM	Rotational speed of the spindle in revolutions per minute.
Anomaly	Boolean indicator signifying whether the data point is considered an anomaly (True/False).

Example Data

Click to expand

time	V_avg_machine	I_avg_machine	kW_machine	kvar_machine	kVA_machine	PF_machine	kWh_machine	kvarh_machine	kVAh_machine	V_avg_spindle	I_avg_spindle	kW_spindle	kvar_spindle	kVA_spindle	PF_spindle	kWh_spindle	kvarh_spindle	kVAh_spindle	RPM	Anomaly
2023-05-04 21:06:25	222.5558624268	3.978289604	1.184034824	0.9684904218	1.529677153	0.774042308	1.816337466	1.503314137	2.358232737	360.457244873	5.533977985	0.152448997	2.475371122	2.480061054	0.061469857	0.164464176	1.580153108	1.598300338	2500	False
2023-05-04 21:06:26	222.5959472656	3.980161905	1.184236288	0.9688802361	1.530079126	0.773969958	1.816667059	1.503583074	2.358657598	360.457244873	5.545718193	0.152648449	2.477063417	2.481761694	0.061508474	0.164548889	1.581528783	1.599682927	2500	False
2023-05-04 21:06:27	222.6310272217	3.984233379	1.184538126	0.9686712027	1.530180931	0.774116392	1.816996574	1.503852010	2.359082460	360.457244873	5.550326347	0.152649045	2.478666782	2.48336339	0.061468504	0.164590508	1.582217336	1.600376129	2500	False

Usage

To use this project, follow these steps:

1. Set Up Your Environment: Ensure you have followed the installation instructions to set up your environment.

2. Run the Streamlit Application: Open a terminal and run the following command to start the Streamlit application:

   streamlit run main.py

3. Interact with the Web Interface: Follow the instructions on the web interface to use the application:

   • Upload Your Dataset: Upload your dataset file in CSV format using the file uploader in the configuration panel on the left side of the interface.

   • Configure Parameters: Set up the necessary parameters in the configuration panel on the left side of the interface.

   • Execute the Pipeline: Click the 🚀 Execute All button to run the entire pipeline, including preprocessing, training, and evaluation.

   • View the Results: The results, including various plots and evaluations, will be displayed on the right side of the interface.

Configuration

The configuration settings for running the scripts are defined as follows:

Configuration	Description
Model	The machine learning model to be used for prediction.
Sampler	Sampling method to balance the dataset.
Sampling Strategy	Desired number of samples for each class after resampling. You can specify the exact number of samples for non-anomalous and anomalous classes.
Future Steps	Number of future steps to predict. For example, if set to 3, the model will predict steps $K$, $K+1$, and $K+2$.
Window Size	Size of the sliding window for feature extraction. For example, a window size of 3 means using the past 3 time steps $K-3$, $K-2$, $K-1$ to predict the next step.
Test Size	Number of samples to include in the test split. This determines the amount of data reserved for testing the model.
Seed	Random seed for reproducibility. Setting this ensures that the results are consistent across different runs.
Correlation Threshold	Correlation threshold for feature selection. Features with correlation higher than this threshold will be removed to avoid multicollinearity.

Contributing

Contributions are welcome! Please follow these steps to contribute:

1. Fork the repository.
2. Create a new branch (git checkout -b feature-branch).
3. Commit your changes (git commit -am 'Add new feature').
4. Push to the branch (git push origin feature-branch).
5. Create a new Pull Request.

License

This project is licensed under the MIT License. See the LICENSE file for details.

Contact

For any questions or inquiries, please contact m11207330@mail.ntust.edu.tw