Demand Forecasting Based on Machine Learning for Mass Customization in Smart Manufacturing

Abstract

Mass customization is essential for smart manufacturing. In particular, generating demand forecast is undoubtedly the most important part of any industry. Appropriate demand forecasts make S&OP quality, which greatly contributes to overall corporate management. In addition, proper stock can be maintained to save the costs of maintaining multiple warehouses. In this paper, we find out why mass customization is needed in smart manufacturing and find appropriate demand forecasting techniques by comparing the traditional time series technique ARIMA analysis with the nonlinear network model. Afterwards, the company develops an algorithm to evaluate the sales process by finalizing the production plan by evaluating the expected inventory through mathematical modelling.