Search Results

It is known that the collisions caused by lane departure events account for range of percentages among the countries studied. To help prevent such collisions, the Lane Departure Warning (LDW) system has started to be introduced in production vehicles, but there is little research on its benefits and limitations so far. In this paper we performed an in-depth analysis of the collisions and driver-related essential variables for the lane-departure collision scenarios and demonstrated the benefit estimation process. The benefit of the LDW system is estimated by comparing lane departure events when the vehicle has no LDW, and how they change with the addition of LDW. The event without LDW was modeled in 5 phases: (1) before departure, (2) starting of the departure, (3) departed the lane, (4) at the impact with an object, and, (5) after the impact. “An extensive analysis was conducted of traffic crash data compiled by the Institute for Traffic Accident Research and Data Analysis (ITARDA).

A number of active safety systems are already developed to support drivers’ decision and action to help avoid accidents, but further enhancement of those active safety systems cannot be accomplished without increasing our understanding on driver behaviors and their interaction with vehicle systems. For this reason, a state-of-art driving simulator (DS) has been developed that creates very realistic scenarios as a means of realizing these requirements. The DS consists of a simulator cabin, turntable (inside the dome), a 6-DOF hexapod system, shakers (vehicle vertical vibration actuators), and a motion system capable of moving 35 meters longitudinally and 20 meters laterally. The system is also capable of projecting images of actual city streets and highways onto a 360° spherical screen inside of the dome. As a result, the DS is able to reproduce a driving environment that is very similar to real driving.

Naturalistic driving data has been accumulated by driving data recorders to understand factors that contribute to collisions. Among the rear end conflicts at signalized intersections in the data, conflict data between the following vehicles and suddenly stopping lead vehicles were frequently observed just after their start. To investigate the following drivers' behavior in a realistic driving situation without collision danger, an instrumented vehicle equipped with a liquid-crystal display ahead of the windshield was developed, and an experiment reproducing such conflict on the display was conducted. It was found that a lead vehicle's rapid start (2.8 m/s₂ on average) before quitting its right turn caused the following vehicle's brake reaction time to be longer than a slow start (0.8 m/s₂ on average) did. This result suggests that a following driver's premature decision to start rapidly increases the risk of rear end collisions.

Pre-collision system (PCS) that basically consists of warning buzzer, brake-assist and automatic brake functions is designed to help mitigate injury occurring in frontal impacts where preceding vehicles are impacted from the rear. Since the benefit of each function is influenced by drivers' reaction before collision, it is difficult to estimate quantitative effectiveness for improvement. This paper proposes an approach to establish a Safety Impact Methodology (SIM) for the effectiveness estimation of a PCS by utilizing driving simulator (DS) test results and Event Data Recorder (EDR) data. The estimation procedure consists of four steps. Firstly, the PCS functions were modeled as linear transfer functions with gain, delay, etc. Test vehicle's properties were measured and modeled as an example. Secondly, the transfer functions were applied to each DS test result, and the speed reduction from travel speed to impact speed was calculated by the model assuming that PCS was installed.