Technological advances have led to the development of autonomous driving. The SAE (Society of Automotive Engineers) defined five automation levels in order to differentiate the responsibilities between the driver and an automated driving system, ranging from “No Automation” (L0), to “Conditional Automation” (L3) and “Full Automation” (L5). The transition to full automation, however, brings new risks, such as mode confusion, overreliance, reduced situational awareness and misuse. Therefore, led by SWOV and together with many other parties, a 4-year project MEDIATOR is launched. The goal is to create a self-learning mediating system which guarantees safe, real-time transition of control between human driver and automated system depending on who is better fit for drive (Mediator, 2019).When the automation system reaches its operational limit in a given traffic situation, the automation issues a so-called take-over request (TOR), asking the driver to take back control of the vehicle (Gasser & Westhoff, 2012; Hoeger et al., 2011). The Human Machine Interface (HMI) plays a critical role on passing signal from the automated vehicle to driver when TOR happens. Therefore, it is of vital importance in avoiding misunderstandings, misuse, over-reliance, reduced situational awareness and mode confusion. However, nowadays, there are various different HMI design for autonomous vehicles in the market. Develop an unified HMI design principle to regulate the autonomous vehicle industry is also one of the tasks of MEDIATOR project.The graduation project is focusing on the HMI research and design when takeover happens ( see Figure 1). The key research questions of the graduation project are “when is it needed to communicate what kind of information and how to communicate the information with the driver during takeover?” So exploring the scenarios of TOR and then define the similarities and differences are one of the tasks. Simultaneously, from a research of Bazilinskyy and colleagues (2018), the means to communicate take-over requests are divided into three main categories that may be used in different level of urgency for take-over transition: visua l(written messages or signals shown on cluster, head unit or other in-vehicle displays), auditory (sonorous signals or voice messages), and vibrotactile (vibration of the steering wheel or seat). So it will be one of the challenges to explore which modalities to communicate TOR are the most effective means. In addition, driver’s situational awareness (SA can be defined simply as “knowing what is going on around us”) predicts the takeover performance. Drivers are better able to respond to hazards when they’re aware of the driving context. Therefore, enhancing the SA is also another challenge for the design of the takeover journey.In conclusion, the objective of the project is to design the HMI of autonomous vehicles in order to enhance situation awareness for a better take over transition/journey.