Artificial Intelligence (AI) use in business: Artificial Neural Network modelling for predicting cost, minimising waste and optimising resource utilisation in furniture industry

The primary objective of this study is to develop a predictive model for estimating the cost of kitchen cabinets prior to the production process by employing artificial neural networks (ANN). The proposed model is structured with five input parameters related to kitchen cabinet specifications, a hidden layer comprising ten neurons, and a single output node representing the predicted cost. The model was trained using the Neural Fitting Tool available in the MATLAB environment. The MATLAB code and the dataset utilised in the study are also provided for reproducibility. Upon completion of the training phase, the model achieved a coefficient of determination (R-value) of 0.9716. Subsequent testing of the model yielded an R-value of 0.9682, indicating a high level of predictive accuracy. Corresponding regression plots are presented and discussed within the study. This approach demonstrates the potential of ANN-based models to improve cost estimation processes in the furniture manufacturing industry. Furthermore, by enabling data-driven decisions prior to production, such models contribute to sustainability efforts by minimising material waste, optimising resource utilisation, and reducing associated carbon emissions. The study also suggests that the methodology can be expanded by training the network with different types of data relevant to other segments of the furniture industry.