There have been a number of digital twin (DT) frameworks proposed for multiple disciplines in recent years. However, there is a need to develop systematic methodologies to improve our ability to produce DT solutions for the nuclear fuel industry considering specific requirements and conditions exclusive to the nuclear fuel manufacturing cycle. A methodology tailored for nuclear fuel production is presented in this paper. Due to the nature of the chemical processes involved in fuel manufacturing, we highlight the importance of using a combination of physics-based and data-driven modelling. We introduce key technologies for DT construction and the technical challenges for DT are discussed. Furthermore, we depict typical application scenarios, such as key stages of the nuclear manufacturing cycle. Finally, a number of technology issues and research questions related to DT and nuclear fuel manufacturing are identified.