Fei Yan, Jürgen Pichen, Sven Schmitz, Kai Sklorz, Martin Baumann

• With the aim of identifying suitable indicators and criteria for evaluating the safe human-machine interaction for SAE level 3 systems up to 60 km/h in the context of automated driving, this research project has started with a focus group interview to identify relevant publication channels and list of keywords regarding indicators for the evaluation of human-machine interaction at SAE Level 3. Based on the identified list of keywords, literature reviews have been conducted to extract relevant publications from the identified publication channels. According to the defined inclusion and exclusion criterion, 38 papers have then been selected and used for meta-analysis to study the influence of different takeover situations on takeover performances. The results of meta-analysis have indicated that drivers’ takeover performances measured by the categories of takeover time, takeover quality and subjective workload are different in static and dynamic situations. After that, expert interviews have been conducted with six international experts to help interpret the results of meta-analysis and develop checklist items. In the end, 16 checklist items assigned in six categories of system requirements have been developed and can be used by international experts to evaluate the safety of the human-machine interaction of SAE Level 3 systems up to 60 km/h in production vehicles. This checklist has been further developed to an online application, which can be used as an easy-to-implement and efficient evaluation procedure in relation to the traffic safety relevant interaction quality of the system.