In order to improve road traffic safety, increasingly sophisticated and robust collision avoidance systems are being developed. When employed in safety-critical situations, however, the interaction between the human factors and these systems may increase the complexity of the task of driving. Due to these human factors, the ability of the driver to respond to various traffic dangers is considered to be a function of the level of automation, balance of control authority, and the innate ability of the driver. For the purpose of this study, a driving experiment was designed using two types of lane change collision avoidance systems. One was a haptic warning system that provides a steering force feedback to avoid hazardous lane change, and the other, a semi-autonomous system that provides an automatic action to prevent hazardous lane change. While drivers had the final authority over the haptic system, they were unable to override the automatic action. Both systems were examined in three conditions: i) hazard that can be detected only by the system, ii) hazard that can be detected only by the driver, and iii) combined hazards. The different support systems were applied to the different hazards resulting in significant differences in drivers’ reaction time and steering behavior. The drivers’ subjective post-hazard assessments were significantly affected by the type of encountered hazard.
MODIFIED HAMILTONIAN ALGORITHM FOR OPTIMAL LANE CHANGE WITH APPLICATION TO COLLISION AVOIDANCE
