The Research and Application of Lane Change Adviser

Abstract

As technology progresses toward NHTSA defined Level 2 driver assistance systems with increasingly advanced sensing capabilities, opportunities arise to enhance driver assistance technologies beyond their current state and assist drivers in the completion of increasingly complex maneuvers. The Lane Change Advisor (LCA) is a feature that utilizes the capabilities of available on-board sensing modalities to infer driver intent to make a lane change and provide contextual environmental cues which when combined with observed driver actions inform a decision framework for the purpose of aiding the driver in maintaining their intended travel speed through the opportunistic identification of lane change maneuver opportunities. The driver intention inferences draw upon forward looking sensor measurements in lane of travel for both a dynamic perceptual model and a pseudo-steady-state traffic following component. The dynamic element utilizes the concept of visual looming, based the human perception of rapid visual image expansion on the retina as the distance between driver and obstacle ahead decrease over time. The steady-state component utilizes commands from Adaptive Cruise Control (ACC) during vehicle following for decision making. This presentation will present the decision framework, developed for a vehicle implementation of this feature as well as present results from on-road vehicle testing.