Background
The @CITY research initiative brought together 15 partners from the automotive industry, supplier industry, software development and universities. Divided into the two projects @CITY and @CITY-AF, concepts, technologies and prototype applications were developed to enable automated driving in complex urban spaces. The aim was to make the urban traffic of the future as safe, comfortable and efficient as possible for all road users. The research initiative was supported by the Federal Ministry of Economics and Climate Protection (BMWK) with a funding volume of around 20 million euros.
The IAD was represented in the @CITY-AF project and conducted research in the sub-project “Human-vehicle interaction”. This sub-project focused on the interaction between the three protagonists: vehicle users, automated vehicles and other road users. The IAD dealt with the question of how everyday human forms of communication in road traffic (eye contact, gestures, etc.) can be “translated” to automated systems without misunderstandings occurring.
Further information on the project can be found at https://www.atcity-online.de/?language=en
Method
- Survey using online questionnaires
- Driving simulator tests (complete vehicle mock-up and self-developed automation controller with driving functions according to SAE L3
- To analyze the effects of non-driving activities during highly automated driving, extensive tests were carried out in the driving simulator with approx. 62 test persons. Various measurement methods were used during the tests:
- Gaze behavior (DIKABLIS eye-tracking system)
- Performance and attention distribution (stimulus-response test according to ISO DIN 17488)
- Stress (ECG, skin conductance)
- Situation awareness (SAGAT – Situation Awareness Global Assessment Technique)
Results
The survey, which was completed by 164 individuals, enabled the identification of the preferred NDRTs for highly automated driving. Subsequently, the five NDRTs identified as relevant were subjected to further examination through a stimulus-response test in order to assess the suitability of these for the subsequent evaluation of the available human resources.
The comprehensive study, which included 62 test subjects, demonstrated that gaze behaviour, mental stress, situational awareness and the capacity to regain control exhibited notable variations contingent on the non-driving activities under investigation. A multiple regression analysis and machine learning algorithms were employed to ascertain the influence of mental stress and situational awareness on the ability to take back control.
