Improving AI-based modelling for mass customisation strategy: Insights from automotive industry

The potential of the mass customisation (MC) strategy is continuously expanding through the implementation of Industry 4.0/5.0technologies.In this study, a novel approach to enhancing the potential of Artificial Intelligencebased techniques for MC level prediction was presented.Firstly, the empirical research was conducted among more than 150 European manufacturing companies in the automotive sector concerning factors influencing the increase in the level of MC, as well as the Industry 4.0 or 5.0 technologies applied.Next, the collected survey data was pre-processed to eliminate infrequent responses and subsequently analysed using the k-means clustering method for labelling purposes.The resulting dataset was then used to fine-tune parameters and train a classifier based on a multilayer perceptron (MLP), aimed at identifying the MC level in automotive enterprises.The developed classifier achieved average accuracies of 0.94 and 0.93 for training and testing, respectively.The proposed approach was validated through real-life case studies from the automotive industry, demonstrating the adaptability of our concept and its potential for diverse industrial applications.