Touch Screen Maneuver Approval Mechanisms for Highly Automated Vehicles: A First Evaluation

Abstract

Prototypes of highly automated vehicles are already able to drive on public roads, however fully automated rides where humans in the vehicle have only the role of a passenger regardless in which environment they travel are far away. Major issues are limited sensor range, mixed traffic, and an insufficient capability of classifying situations. We propose that vehicles can cooperate with the human inside to overcome such system boundaries. A possible input modality for driver-vehicle interaction in such scenarios are touch screens. We investigated three implementations regarding different confirmation processes to avoid erroneous inputs. Our evaluation in a driving simulator with 18 participants indicates that drivers prefer a one-tap selection, however they accept error prevention mechanisms like a confirmation dialog to approve maneuvers in more dynamic and complex scenarios. Moreover, these mechanisms did not have a negative effect on usability and workload.