Carnegie Mellon Robotics Institute
Louis Tijerina, James J. Jackson, Dean Pomerleau, Richard A. Romano, and Andrew D. Pertersen
ITS America Sixth Annual Meeting, April, 1996.
|This report presents the results of a simulator-based exploratory study of Collision Avoidance System (CAS) concepts suitable for roadway departure collision avoidance. Roadway departure crashes account for significant percentages of both the total number of crashes and the number of fatal crashes in the United States. CAS support that could avert or minimize the severity of even a fraction of these crashes would prove beneficial to the driving public.
The purpose of the study was to evaluate the following items from a driver-oriented perspective. Sixty-four volunteers participated at the Iowa Driving Simulator (IDS), a six-degree-of-freedom, moving-base simulator with a wide field-of-view image generation system. Sixteen of the participants were randomly assigned to serve in a control group without CAS support; the remaining 48 participants were randomly assigned to groups of 16 in each of three CAS Interface groups: auditory, haptic, or combined-modality. Within the CAS groups, participants were further assigned to different levels of four factors: directionality of CAS display (directional or nondirectional), Onset (early CAS onset or late CAS onset), and Algorithm (Time-to-Line-Crossing [TLC] versus Time-to-Trajectory Divergence [TTD]) for lanekeeping .
All participants were assigned to either high or low magnitude hazard conditions. The lateral disturbance collision hazard involved a simulated lateral offset (Le,. wind gust) applied while the driver was engaged in an in-vehicle distractor task; low hazard magnitude was equated to a small lateral offset and high hazard magnitude was equated to a large lateral offset. In addition, participant performance was assessed during normal (non-hazard) lanekeeping early and late in a 40-minute simulator session.
Results suggest that the concept of a roadway departure CAS has potential. Given that a CAS is to be developed. the data indicate that directional displays have some performance advantages and consumer preference. Based on the evidence gathered in this study, auditory and haptic interface types merit further investigation and development. However, a combined-modality display may be a source of information overload to a driver. Early onset is also advised for the lateral CAS concept. While it appears that TLC may be a preferred algorithm for a lateral roadway departure CAS, it is associated with somewhat greater driver steering effort. Furthermore, both TLC and early onset are associated with more CAS activations, a potential source of nuisance alarms. Finally, it must be acknowledged that drivers were, on average, lukewarm to the CAS concepts included in the study. While this is perhaps not surprising given the exploratory nature of the research, it suggests that driver acceptance will need to be a key goal of efforts to bring such Intelligent Transportation System (ITS) concepts to fruition. The potential exists for advanced technology to contribute to enhanced highway safety, but the human factors remain crucial for achieving such gains.
Associated Center(s) / Consortia:
Vision and Autonomous Systems Center
Associated Lab(s) / Group(s): NavLab
|Louis Tijerina, James J. Jackson, Dean Pomerleau, Richard A. Romano, and Andrew D. Pertersen, "Driving Simulator Tests of Lane Separture Collision Avoidance Systems," ITS America Sixth Annual Meeting, April, 1996.|
author = "Louis Tijerina and James J. Jackson and Dean Pomerleau and Richard A. Romano and Andrew D. Pertersen",
title = "Driving Simulator Tests of Lane Separture Collision Avoidance Systems",
booktitle = "ITS America Sixth Annual Meeting",
month = "April",
year = "1996",
|The Robotics Institute is part of the School of Computer Science, Carnegie Mellon University.|
Contact Us | Update Instructions