Browse Publications Technical Papers 2015-01-0032

Scenario Analysis of Near-miss Incidents to Enhance Pedestrian Collision Warning System 2015-01-0032

First, this paper focuses on classification of near-miss incidents with pedestrian into typical scenarios, and then incident data are analyzed under the assumption that pedestrian detection system has a pre-determinant sensing performance.
The data of 220 near-miss incidents with pedestrian when own vehicle goes straight are employed. We proposed a set of eight scenarios which structure is not necessarily hierarchical. The eight scenarios cover about 75% of the incidents.
To highlight the characteristics of the eight scenarios, two kinds of vector-diagrams are introduced which illustrate the transition of relative position between pedestrian and own-vehicle, and TTC-TTV (Time To Collision - Time To Vehicle) relation respectively. For example, in a relative position diagram, we can grasp the pedestrian's relative position which the detection system finds the pedestrian first when he/she appeared with pre-determinant wide detection angle. In TTC-TTV diagram, we can observe that the own-vehicle and the pedestrian approach each other maintaining a safe collision gap (difference between TTC and TTV) or not. After the analysis of the scenarios using vector diagram, we also discuss the effectiveness of these scenarios from the aspects of speed profile of own-vehicle, because the car speed is one of the most important factors which characterize near-miss incidents.
From these analyses and discussions using the vector diagrams, the characters of each scenario become apparent. Scenario analysis shown in this paper must contribute to development of pedestrian collision warning system which aims to reduce pedestrian accidents.


Subscribers can view annotate, and download all of SAE's content. Learn More »


Members save up to 18% off list price.
Login to see discount.
Special Offer: Download multiple Technical Papers each year? TechSelect is a cost-effective subscription option to select and download 12-100 full-text Technical Papers per year. Find more information here.