PROJECT TITLE :

An Unsupervised Approach for Inferring Driver Behavior From Naturalistic Driving Data

ABSTRACT:

Intelligent transportation systems are ready to gather large volumes of high-resolution information. The amount of data collected by these systems can quickly overwhelm the ability of human analysts to draw meaningful conclusions from the data, significantly in large-scale multivehicle field trials. As advanced driver assistance systems develop, they will also be needed to form a made and high-level understanding of the planet from the information they receive, as well as the behavior of the driver. These applications encourage the requirement for unsupervised tools capable of forming a high-level summary of low-level driving knowledge. This paper presents an unsupervised methodology for changing naturalistic driving knowledge into high-level behaviors. The proposed technique works in two steps. In the first step, inertial data are automatically decomposed into linear segments. In the second step, the segments are assigned to high-level driving behaviors. The proposed method is computationally economical and completely unsupervised and needs minimal preprocessing. Though the tactic is unsupervised, the clusters made exhibit high-level patterns that may simply be associated with driving behaviors like braking, turning, accelerating, and coasting. The effectiveness of the proposed algorithms is demonstrated in an offline application where the objective is to summarize inertial knowledge into driving behaviors. The strategy is also demonstrated in an online application where the aim is to infer the present driving behavior using solely inertial knowledge. Both experiments were conducted using driving knowledge collected in natural driving conditions.