Digitalizing R&D in manufacturing sector : machine learning, infrastructure, system architecture and knowledge management

Abstract

This thesis addresses the topic of data utilization and data analytics in research and development (R&D) functions of the manufacturing sector. Many companies in the manufacturing sector have generated significant quantities of data in their histories, but only a tiny part of these data is utilized. With the significant progress in big data analytics and machine learning, the companies in the manufacturing sector are able to upgrade their R&D capability by establishing a system to better collect and analyze their data. Using machine learning can tremendously enhance R&D's capability in interpreting data and giving recommendations regarding solutions. The data system could also help improve an R&D organization's productivity by significantly reducing repeated work. This thesis designs an R&D system that collects R&D data by lab automation, analyzes data by built-in machine learning algorithms, and provides recommendations by gathering inputs for development targets. This thesis also covers aspects of knowledge management within the corporation when implementing such a data system. The organizational capability to implement this data system is also discussed.