Machine Learning for Intelligent Industrial Design

Machine learning (ML) techniques have been used to build intelligent software for several domains. This paper reviews and discuss opportunities for using these techniques to build intelligent software for industrial design. Industrial design, sometimes called product design or new product development, is the process of conceiving products to be mass-produced in factories. It consists of several steps such as: analyzing potential customers wants and needs, planning, prototype design, and user evaluation. During each of these steps, data can be collected as documents such as product specifications and feedback forms, or by other means such as using sensors. A promising way of improving these processes to reduce costs (time and investments) and produce better designs, is to analyze data generated or used during product design using ML techniques, and to build intelligent design software. Although several studies have been carried out on this topic, there remains numerous research opportunities. This paper provides a survey of recent studies related to the use of ML in industrial design. The goal is to provide an introduction to this emerging research area and highlight limitations of previous studies and opportunities.