Developing complex crash warning simulations for human factors evaluations (original) (raw)
Related papers
Using driving simulators to assess driving safety
Accident Analysis & Prevention, 2010
Changes in drivers, vehicles, and roadways pose substantial challenges to the transportation safety community. Crash records and naturalistic driving data are useful for examining the influence of past or existing technology on drivers, and the associations between risk factors and crashes. However, they are limited because causation cannot be established and technology not yet installed in production vehicles cannot be assessed. Driving simulators have become an increasingly widespread tool to understand evolving and novel technologies. The ability to manipulate independent variables in a randomized, controlled setting also provides the added benefit of identifying causal links. This paper introduces a special issue on simulator-based safety studies. The special issue comprises 25 papers that demonstrate the use of driving simulators to address pressing transportation safety problems and includes topics as diverse as neurological dysfunction, work zone design, and driver distraction.
Driving Simulation for Crash Avoidance Warning Evaluation
2000
This paper describes the development of a fully-integrated, medium-performance, fixed-base driving simulator for testing and evaluating crash avoidance warning systems. Three major aspects of the ongoing effort are described: our motivation for constructing the simulator, the system architecture, and the crash avoidance warning systems. The paper concludes with an assessment of our progress to date. The physical simulator is a
2011
The Crash Warning Interface Metrics (CWIM) project addressed issues of the driver-vehicle interface (DVI) for Advanced Crash Warning Systems (ACWS). The focus was on identifying the effects of certain warning system features (e.g., warning modality) and on establishing common methods and metrics that may be generally applied for evaluating DVIs in different vehicles. The project did not have the goal of proposing standard interfaces for particular warning functions, but it did consider implications for design. The project included analytical activities and five experiments. Each experiment investigated the effects of ACWS DVI on driver behavior or comprehension using a different methodology. An objective of these studies was to determine the appropriateness of the various methodologies for use in subsequent human factors research on ACWS DVIs. Implications were discussed for methods to evaluate DVIs including driving scenarios, research participant characteristics, pre-familiarizati...
Validating a driving simulator using surrogate safety measures
Accident Analysis & Prevention, 2008
Traffic crash statistics and previous research have shown an increased risk of traffic crashes at signalized intersections. How to diagnose safety problems and develop effective countermeasures to reduce crash rate at intersections is a key task for traffic engineers and researchers. This study aims at investigating whether the driving simulator can be used as a valid tool to assess traffic safety at signalized intersections. In support of the research objective, this simulator validity study was conducted from two perspectives, a traffic parameter (speed) and a safety parameter (crash history). A signalized intersection with as many important features (including roadway geometries, traffic control devices, intersection surroundings, and buildings) was replicated into a high-fidelity driving simulator. A driving simulator experiment with eight scenarios at the intersection were conducted to determine if the subjects' speed behavior and traffic risk patterns in the driving simulator were similar to what were found at the real intersection. The experiment results showed that speed data observed from the field and in the simulator experiment both follow normal distributions and have equal means for each intersection approach, which validated the driving simulator in absolute terms. Furthermore, this study used an innovative approach of using surrogate safety measures from the simulator to contrast with the crash analysis for the field data. The simulator experiment results indicated that compared to the right-turn lane with the low rear-end crash history record (2 crashes), subjects showed a series of more risky behaviors at the right-turn lane with the high rear-end crash history record (16 crashes), including higher deceleration rate (1.80 ± 1.20 m/s 2 versus 0.80 ± 0.65 m/s 2), higher non-stop right-turn rate on red (81.67% versus 57.63%), higher right-turn speed as stop line (18.38 ± 8.90 km/h versus 14.68 ± 6.04 km/h), shorter following distance (30.19 ± 13.43 m versus 35.58 ± 13.41 m), and higher rear-end probability (9/59 = 0.153 versus 2/60 = 0.033). Therefore, the relative validity of driving simulator was well established for the traffic safety studies at signalized intersections.
International Journal of Automotive Engineering
Automated driving safety evaluation predominantly relies on scenario-based approaches. In this study, the authors adopt a functional scenario catalogue initially conceived by JAMA to evaluate automated driving safety on limited access highways. The potential of this catalogue to cover real-world crashes was investigated by comparing each scenario in the catalogue with crash patterns from two international data sources: the 2007 NHTSA's pre-crash scenario typology for crash avoidance research report, and the 2020 IGLAD's codebook. The results indicate the potential of the scenario catalogue to comprehensively cover both the NHTSA and the IGLAD crash scenario typologies.
A Conceptual Approach for Using the UCF Driving Simulator as a Test Bed for High Risk Locations
The main objective of this paper is to illustrate the use of the UCF Driving Simulator as a test bed for high-risk locations such as signalized intersections and toll plazas. The Alafaya (SR434) -Colonial (SR50) signalized intersection in Orange County, Florida, is used to validate the simulator. The SR 434 -SR 50 intersection is four-leg 6×5 signalized intersection, and has two left turn lanes and one right turn lane for every approach. Both safety and traffic measures are being used in our effort to validate the simulator and therefore prove its usability as a test bed. Safety will involve identifying the potential safety problems from the crash reports, then we will observe whether subjects will face the same problems in the simulator, thus validating it. The traffic measure would be the speed distribution. Speed measures in the field will be compared with the driving speeds of the subjects crossing the intersection in the simulator, again for validating it. Multiple scenarios is being developed covering the different safety problems and approaches to the intersection. Using the Florida Department of Transportation Crash Analysis Reporting (CAR) System, the crash reports at this intersection for the years 1999 to 2002 were obtained and analyzed. Of the 164 crashes that happened during the 4 years at the intersection, 95 have been rear-end. It has been found that most of the rear-end crashes took place in the right-turn lanes of SR 434. The new approach here is the use of crash history to identify the high risk locations and type of collisions at the intersection, and then observe the "near misses" in the simulation. If both coincide, this could be one of the validation criteria that could be used. A between-subjects 3X2 (3 age groups and 2 genders) factorial experimental design has been modeled. Sixty subjects will drive the simulator, running through 11 scenarios, mainly the right turn lanes on SR434 in both the directions, when the signal is green/red/amber for traffic on SR50 (for traffic conflict validation), the through lanes on SR434 when the signal is green (for speed validation), and the through lane on SR50 when the phasing changes from Green to Amber. Measurements of Speed, Deceleration rate, the position of stop and the steering performance, will be collected. Also the effect of gender and age will be assessed.
An Integrated Driver-Vehicle-Environment (I-DVE) model to assess crash risks
2000
A wide range of driver and vehicle models have been proposed by traffic psychologists, engineers and traffic simulation researchers to assess crash risks. However, existing approaches are often confined within a single discipline and lack concepts that formally express the complexity of interactions between the driver, vehicle and environment as well as the broader scope and the interdisciplinary nature of