EURO 2025 Leeds
Abstract Submission

2812. Machine Learning Based Hierarchical Algorithms: An Application In Plastic Injection Industry

Invited abstract in session TA-28: AI and Machine Learning for Decision Support, stream Decision Support Systems.

Tuesday, 8:30-10:00
Room: Maurice Keyworth 1.03

Authors (first author is the speaker)

1. Alara Baysal
Data Science, Ozyegin University
2. Cevdet Utku Şafak
Industrial Engineering, Ozyegin University
3. Ahmet Eralp Avunduk
Data Science, Özyeğin University
4. Eyyüb Batuhan Ünal
Data Science, Ozyegin University
5. Erinc Albey
Industrial Engineering, Ozyegin University

Abstract

Many plastic injection moulding facilities rely on labor-intensive quality control methods, presenting an opportunity for machine learning algorithms to enhance real-time detection of defective products. This study introduces a hierarchical algorithm that utilizes machine parameters and cavity sensor data to evaluate plastic injection parts' quality and fault types.

We explore the effects of sensor placements, including pressure and temperature sensors near mould gates and cavity edges, and extract features using linear fitting techniques. These features, combined with machine parameters, facilitate the prediction of multiple fault types. The hierarchical approach provides valuable insights into process deviations by capturing interactions between machine settings and sensor data.

The results show the algorithm achieves approximately 90% accuracy in 1/0 fault detection and between 90% and 95% for various common faults in the industry. These findings could help recommend real-time corrective actions, leading to improved equipment efficiency and reduced downtime.

Keywords

Status: accepted


Back to the list of papers