Online adaptation of the Level of Haptic Authority in a lane keeping system considering the driver’s state

Abstract

Driver distraction is an important factor of accidents. Not only in manual driving, the driver state information is also interesting to consider in automated driving systems in order to provide the driver with a suitable assistance level in respect to his evolving needs. The concepts of adaptive automation and human machine cooperation suggest that the authority of the automation should be adapted in real time according to the situation. However, contrary to existing studies that demonstrated the benefits for continuous fixed haptic feedback in the lane keeping task, evidence regarding the potential benefits of online adaptation of the level of haptic feedback is still lacking. In this framework a study is conducted in order to investigate the effects of online adjusting the authority level of a lane keeping assist system to match the driver’s distraction state while engaging in a demanding secondary task. The study took place in the SHERPA-lamih driving simulator. A comparison has been made between manual driving asa baseline, a Lane keeping Assist (LKA) providing a fixed and continuous haptic feedback and an Adaptive Lane Keeping Assist (ALKA). The analysis accounted for the driving performance and effort along with the subjective ratings of comfort, safety, control and workload. The results were consistent with the previous studies that showed the benefits of fixed haptic feedback under normal driving conditions. Moreover, the study established the benefits of adapting the level of haptic authority when the drivers were engaged in a secondary task. Furthermore, some design issues are highlighted for the design of effective adaptive automation.