Machine learning approach for predicting production delays: a quarry company case study

Area of Science:

• Industrial Engineering
• Data Science
• Machine Learning

Background:

• Industry 4.0 necessitates a shift from reactive to predictive manufacturing.
• Implementing predictive maintenance presents challenges for industrial businesses.
• Big data analytics and machine learning are key enablers for Industry 4.0.

Purpose of the Study:

• To demonstrate the application of data analytics and machine learning for predicting production delays.
• To provide a case study in a quarry firm for predictive modeling.
• To enhance decision-making processes in manufacturing environments.

Main Methods:

• Utilized the Cross Industry Standard Process for Data Mining (CRISP-DM) methodology.
• Developed five predictive models using algorithms: Decision Tree, Neural Network, Random Forest, Nave Bayes, and Logistic Regression.
• Analyzed a dataset of six months of production records from two machines.

Main Results:

• Multilayer Perceptron Neural Network and Logistic Regression models showed superior performance.
• Accurate prediction of production delays achieved with a F-measure score of 0.973.
• Identified key machine learning algorithms for effective production delay prediction.

Conclusions:

• Machine learning and data analytics effectively predict production delays in industrial settings.
• The study validates the value of predictive modeling for optimizing quarry operations.
• Improved decision-making through accurate delay prediction leads to reduced production line interruptions.