Citation

Naujoks F, Purucker C, Wiedemann K, Neukum A, Wolter S, Steiger R. Accid. Anal. Prev. 2017; 108: 147-162.

Affiliation

Ford Werke GmbH, Spessartstrasse Tor 54, 50725 Cologne, Germany. Electronic address: rsteige3@ford.com.

Copyright

(Copyright © 2017, Elsevier Publishing)

DOI

10.1016/j.aap.2017.08.027

PMID

28886450

Abstract

This study investigated driver performance during system limits of partially automated driving. Using a motion-based driving simulator, drivers encountered different situations in which a partially automated vehicle could no longer safely keep the lateral guidance. Drivers were distracted by a non-driving related task on a touch display or driving without an additional secondary task. While driving in partially automated mode drivers could either take their hands off the steering wheel for only a short period of time (10s, so-called 'Hands-on' variant) or for an extended period of time (120s, so-called 'Hands-off' variant). When the system limit was reached (e.g., when entering a work zone with temporary lines), the lateral vehicle control by the automation was suddenly discontinued and a take-over request was issued to the drivers. Regardless of the hands-off interval and the availability of a secondary task, all drivers managed the transition to manual driving safely. No lane exceedances were observed and the situations were rated as 'harmless' by the drivers. The lack of difference between the hands-off intervals can be partly attributed to the fact that most of the drivers kept contact to the steering wheel, even in the hands-off condition. Although all drivers were able to control the system limits, most of them could not explain why exactly the take-over request was issued. The average helpfulness of the take-over request was rated on an intermediate level. Consequently, providing drivers with information about the reason for a system limit can be recommended.

Copyright © 2017 Elsevier Ltd. All rights reserved.

Language: en

Keywords

Automated driving; Controllability; Lateral scenarios; Partial automation