F2008-02-029
Safe Driving Education through Simulations Based on Actual Driving Data When Entering a Non-Signalized Intersection
In our previous study, we observed that many drivers without the right of way often exhibit potential dangerous driving behaviors when entering a non-signalized intersection. We considered that such drivers have incorrect awareness of careful driving and this causes dangerous driving behaviors. We identified the incorrect awareness as a set of inner factors. In order to prevent right-angle collisions at non-signalized intersections, drivers need to notice their inner factors that lead to their dangerous driving behaviors. For this purpose, we defined five elemental driving behaviors that characterize a series of driving behaviors when entering a non-signalized intersection. We also identified the inner factors that create errors in each elemental driving behavior. We called the relationships between the elemental driving behavior and inner factors the "inner factor model." Next, we made algorithms that determine the time-series of a carīs speed and a driverīs eye movement by connecting the five elemental driving behaviors. These algorithms are used in simulation based on actual driving behavior. The simulation program generates various dangerous situations by changing the speed and timing of crossing cars and bicycles. Simulation results show how dangerous the driverīs driving behaviors are. As a case study, we evaluated driving behaviors of a skilled driver and an unskilled driver and confirmed the effectiveness of the simulation system.
In this study, we examined the driving behaviors of various kinds of drivers by simulation. We used combinations of the elemental driving behaviors that can be recorded in actual driving on the road. We also estimated the state of inner factors by using the inner factor model and identified the relationships between dangerous driving behaviors and the inner factors that the driver actually has. Through simulation using such data, we can collect a lot of data on the relationships between the patterns of a driverīs inner factors or the　elemental driving behaviors and the possibility of causing right-angle collisions or close calls. As an example of the results, we found that drivers who donīt sufficiently slow down and make safety checks around the stop-line cannot have sufficient time to make safety checks when entering an intersection. We also found that they have a strong possibility of getting into a right-angle collision. This is due to their inner factors of having little awareness of bicycles and pedestrians around the stop-line. Such drivers cannot take a proper action toward not only bicycles but also crossing cars that appear from the right side. Thus, by conducting simulations under various conditions, we can suggest how driving behaviors caused by the inner factors are dangerous. This enables us to educate unskilled drivers for fundamentally improving dangerous driving behaviors by showing accidents caused by their inner factors. Through this approach, we can make a guideline for each unskilled driver based on his/her actual driving record on the road.
Session: MMI - Tasks and Assistance