Assistance Method for Merging by Increasing Clarity of Decision Making

Merging into the flow of high-speed traffic on roadways, such as expressways, is a demanding driving task for nonexpert drivers. The difficulty of merging is thought to arise from the difficulty of judging the location where the drivers should merge. In this paper, we propose a mathematical model of driver decisionmaking that can predict the location where the driver will merge. In addition, we demonstrate that the entropy of the judgment introduced based on the proposed model has the ability to describe the differences in skills of judgment in merging position as well as the difficulty of decision-making in a given situation. Furthermore, we propose a driver assistance method to decrease the difficulty of decision-making with respect to the merging position during the preparation phase based on the proposed decision-making model when the difficulty is judged as being higher than a given threshold. The proposed method recommends acceleration or deceleration to increase the clarity of judgment. Driving simulator experiments demonstrate that the proposed method reduces the difficulty of decision-making for merging and increases the ease of judgment at the early stages of merging preparation. In addition, the proposed method decreases mental demand without leading to an increase in the overall workload, that is, the weighted workload of NASA-TLX.