Safety Assessment of Ecodriving Vehicles on Following Traffic

Abstract

Ecodriving aims to achieve the best fuel efficiency by guiding vehicles to travel at a planned speed trajectory, which is developed by making use of topography or signal phasing and timing (SPaT) information. Previous work has mainly focused on algorithm development or the assessment of ecodriving efficiency. However, there has been little research regarding the safety issues of ecodriving. This study addresses safety concerns for normal driving vehicles (FNV) following ecodriving vehicles (EV). Time to collision (TTC) was used as the dependent variable to quantify the critical situation. The effects on TTC of three network-related factors: initial signal status, ambient vehicles, and speed limit, are considered and investigated. Six testing scenarios (experimental group) where FNV (subjects) followed an EV were generated with JMP (statistical software), and six baseline scenarios with conventional cars only were used as the control group. The driving simulation study was designed and implemented with a drive safety DNS-250 simulator. Twenty-nine volunteer drivers participated in the driving simulation study. The results show that the initial signal status has a significant influence on TTC of FNV following EV in the initial signal part, deceleration part, and overall optimization horizon. In addition, the speed limits have a significant impact the TTC of FNV following EV in the acceleration part. Overall, following an EV is safer than following a non-EV, with the exception of FNV in the initial signal part and deceleration part when the initial signal is green.