Researchers in Japan, Europe, and the United States of America are investigating ways to help reduce pedestrian accidents. Methods of how to evaluate the pre-collision systems (PCS) for pedestrians are being considered with the goal of global dissemination and enhancement of the performance. This paper analyzes accident and near-miss incident data and proposes evaluation conditions for a pedestrian PCS. The development of a test apparatus for assessing the performance of the pedestrian PCS under the proposed evaluation conditions is also described.

First, accident data was analyzed to determine the evaluation scenario. The frequency of each combination of vehicle and pedestrian behavior in pedestrian accidents was investigated. According to the analysis results, the most frequent accident scenario was a pedestrian crossing a road while a vehicle goes straight ahead. This scenario was selected for the evaluation. After collating the accident data in terms of pedestrian age, the research focused on two accident patterns:​ one involving elderly pedestrians and one involving children. The accident scenario evaluation conditions include the position lateral to the vehicle at which the pedestrian appears, the walking direction, vehicle speed, and the like. These specific conditions were set by analyzing pedestrian accident data and near-miss incident data. For accidents involving elderly pedestrians, two evaluation conditions were set:​ crossing from the left during daytime and crossing from the right at night. For children, the evaluation conditions featured a child emerging suddenly from behind a parked vehicle.

Next, a pedestrian dummy capable of evaluating the PCS based on these conditions were developed. The pedestrian dummy must be compatible with the use of automotive millimeter wave radar or cameras by the PCS under the evaluation conditions. For a PCS that uses millimeter wave radar, a pedestrian dummy will be required to have low reflection intensity, a capability to reflect radar waves from the entire body, and a walking motion. For a PCS that uses cameras, a dummy must be capable of simulating a human body with both arm and leg movement. To achieve these requirements, the skeletons of the pedestrian dummy were manufactured from vinyl chloride pipes. The reflection intensity was adjusted by winding a metal tape around the entire body of the dummy. A walking mechanism which moves both the arms and the legs was provided from above the pedestrian dummy.

The developed dummy successfully simulated a pedestrian by achieving a reflection intensity which is virtually identical to an actual pedestrian, and a highly realistic walking motion. Furthermore, the test apparatus was developed to assess the PCS under the proposed evaluation conditions.