An Evaluation Framework to Measure the Performance of ADAS

Yunlong An,Xindi Tang,Xi Lin,Jingyue Zheng,Fang He,Liang Ma,Meng Li

DOI: https://doi.org/10.1061/9780784482292.322

2019-01-01

Abstract:Advanced driving assistant systems (ADAS) which help drivers have become major equipment in vehicles nowadays. However, the functions and performances of ADAS between different manufacturers exhibit large variations, and some unsatisfactory functions may leave drivers bad experiences, even endangering their safety. This study is devoted to establishing an evaluation framework of ADAS considering multiple perspectives, including drivers, vehicles, and the environment by combing objective and subjective factors. The evaluation framework covers the macroscopic factors such as traffic throughput and travel time, and the microscopic measures such as reacting time and drivers' stress. We select adaptive cruise control (ACC) as an example to establish a testing routine. Two typical scenarios are designed, and a real-car experiment at a testing site using three car types is conducted, which is among the first in China. Multi-source data collected from the test were used to demonstrate the effectiveness of the proposed evaluation framework.