HFauto

Human behaviour during highly automated driving

Work Package 1  |  Status: Active
Involved researchers: Daniel Heikoop, Zhenji Lu

In this work package, Daniel and Zhenji will investigate human behaviour when driving with high levels of automation.

In driving simulator experiments, volunteer test drivers will be exposed to scenarios where they are on-board a virtual highly automated vehicle. This vehicle will drive automatically in the vast majority of conditions, and will inform the operator in critical situations, or may request the operator to take over responsibilities. Scenarios will include long distance highway driving as well as metropolitan highway networks, with mixed traffic consisting of simulated automated and manually driven vehicles. The research will focus on human behaviour when driving with automation, will compare this to manual driving, and will quantify effects of automation on drivers of other vehicles driving manually. The virtual highly-automated vehicles will be implemented in cooperation with WP2 (providing the HMI) and WP3 (providing a driver state monitor).

Daniel will study driver awareness and arousal during long monotonous monitoring of automation while being transported in a platoon. He will assess driver states such as vigilance, situation awareness, trust, locus of control, and satisfaction/acceptance. In addition, Daniel will investigate various concepts allowing the driver to engage in secondary activities such as reading or using in-vehicle entertainment. It will be investigated whether and how we can ensure that drivers resume control when needed. Close contact will be maintained with the researchers of WP3 who develop an innovative driver state monitor.

Zhenji will investigate human behaviour during transient manoeuvres, such as merging, splitting, platoon entry, platoon exit, and authority transitions between manual and automated driving. Human responses to unexpected situations, such as sensor failure and computer failure, will be studied as well. Furthermore, Zhenji will investigate if and how often operators disengage the system. Finally, Zhenji will also empirically study the willingness of participants to serve collective versus personal interests (e.g., creating a gap or not for another vehicle).