ROLE OF ARTIFICIAL INTELLIGENCE IN AUTOMATION AND IMPROVEMENT OF EFFICIENCY IN INJECTION MOULDING PROCESSES

Abstract

Most of the usable plastics produced across the world involvethe injection moulding process. Improving the efficiency of the injection moulding process can reduce wastages and improve efficiency without affecting the profitability of a manufacturing plant. One of the most important technologies to improve injection moulding efficiency is through Artificial Intelligence (AI). When combining sensors for data acquisition and AI for automation, the improvement opportunities are endless. In order to understand the trends, opportunities, and challenges in AI integration, we designed a narrative-style review in this study. The study uses peer-reviewed research papers that implemented or compared any machine learning algorithm after 2016 are considered for the study. The review focuses on AI models capabilities in process automation, parameter calculation, quality assurance and sustainability in injection moulding manufacturing process. The review mainly distinguished AI models into four categories: supervised learning, unsupervised learning, deep learning, and reinforcement learning algorithm. The review found that most of the research in AI implementation in injection moulding considers either supervised learning algorithms or deep learning algorithms. Many of the implementations are careful in using reinforcement learning as they don’t want to include trial and error-based training. But supervised learning like Genetic Algorithm and Particle Swarm Optimization (PSO) have better real-world implementation capability compared to other models. The review highlights the problems and opportunities in existing AI implementations in injection moulding.