Role of Artificial Intelligence, Automation, and Machine Learning in Sustainable Plastics Packaging markets: Progress, Trends, and Directions

Abstract

The optimisation of manufacturing processes in terms of resource consumption, waste minimization, and pollutant emissions is gaining prominence, especially in small and medium-sized enterprises (SMEs). The advent of digital technology and the subsequent explosion in data volume is another key factor. There is great potential in the data collected from a wide variety of devices and systems, which is why even smaller businesses require access to clever, dynamic analytic models. Sustainable packaging solutions are gaining significance as the world struggles to address global environmental issues. These solutions are being developed with a growing contribution from artificial intelligence (AI). Artificial intelligence is being used to create sustainable and environmentally friendly packaging. Artificial intelligence (AI)-driven technology can be used, for instance, to determine which packing materials and designs are optimal for a given product. Artificial intelligence can also be used to determine the best methods of packaging, such as those that make the most of recyclable materials or that optimise packaging lines. The term "plastic production automation" refers to the use of automated systems and machines in the production of plastic goods. Computer-aided design (CAD), robots, and other cutting-edge technologies are used to automate and optimise production. In this article, we describe the findings of a study that aimed to determine whether or not small and medium-sized enterprises (SMEs) in the plastics processing industry may benefit from the use of machine learning methods in order to optimise energy consumption and reduce the number of wrongly made plastic parts. The machine data in a plastics manufacturing facility for the automobile industry were recorded and analysed in terms of the material and energy fluxes for this purpose. To find areas for optimisation, these data were trained using machine learning techniques. The project also sought to solve the challenge of analysing manufacturing processes with significant non-linearities and time-invariant behaviour by employing Big Data techniques and self-learning controls. Machine learning can help with this if there is enough data to train the system.