|Highway Driving Simulator||Field Research Vehicles||Sign Design a nd Research Facility||MiniSimTM Driving Simulator||Virtual Reality (VR) Lab|
The Highway Driving Simulator (HDS) is a research tool used in the Human Factors Laboratory for a variety of behavioral studies related to safety and operations conducted for Federal Highway Administration (FHWA) and other stakeholders. The simulator consists of a full automobile chassis surrounded by a semicircular projection screen. Three high-definition projectors render a seamless 200-degree view (motorists' field of view) of high-fidelity, computer-generated roadway scenes. A virtual 360-degree field of view is generated by three liquid-crystal display (LCD) panels used in place of the vehicle's three rearview mirrors.
Located at FHWA’s research center in McLean, VA, the driving simulator is a high-fidelity, state-of-the-art research tool that the Human Factors Laboratory uses to generate multiple driving scenarios for evaluation and analysis.
In 2012, the simulator was upgraded from three degrees of freedom to six degrees of freedom to enhance the motion base. This improvement makes the motion and vestibular perception (the perception of body position and movement) much more realistic for drivers. In addition, the driving simulator has a 120-hertz (Hz) eye-tracking capability (that is, it takes 120 samples per second), which allows researchers to investigate where participants are looking when they drive through various roadway scenarios.
When FHWA introduced the double crossover diamond interchange design in the United States in 2004 in Springfield, MO, the driving simulator played an important role in testing human factors issues related to that interchange. Although France has used the double crossover diamond interchange successfully for 30 years, this freeway interchange design was new to the United States. To assist with U.S. development, Michel Labrousse, director of the Centre d'Etudes Techniques de l'Equipment Normandie-Centre, provided records, signal layouts, and traffic flow and crash data from a groundbreaking installation in Versailles, France.
Many conventional interchanges in urban areas are congested and experience high crash rates. In comparison to a conventional diamond interchange, a double crossover diamond design involves drivers crossing from the right side of the road to the left side and then back, thus combining left-turning and through traffic movements. Because of this new design, one human factors concern was that drivers might become confused and make a dangerous maneuver. To evaluate this concern, FHWA researchers created visualizations in the simulator of various driving scenarios.
This screenshot image from FHWA's driving simulator shows a sample scenario used in a human factors study of the diverging diamond interchange.
The Missouri Department of Transportation (MoDOT) designed and built the Nation’s first double crossover diamond interchange in Springfield, MO, and opened it to traffic in June 2009. During the design phase, the Missouri engineers visited the Human Factors Laboratory to virtually drive through a simulated double crossover diamond interchange. At the same time, the laboratory’s researchers provided feedback on the details of the MoDOT design. The visualization and testing in the driving simulator helped to alleviate safety concerns about the new design. The FHWA researchers then created video clips from the simulation scenarios to facilitate outreach to the Missouri public.
A current study using the simulator examines issues related to driver distraction. Researchers are investigating whether advertising on changeable message signs is distracting to drivers. Some of the measures used in the study include the number and duration of eye glances to each sign, and whether participants notice a sign telling them to exit the freeway because there is a crash ahead. The researchers also want to determine whether there is any correlation between potential distraction from advertising on changeable message signs and safety concerns. The study is in the data-collection phase, and the results are expected to take another year.
A field research vehicle (FRV), an instrumented 2007 sport utility vehicle (SUV), is another tool in use at the Human Factors Laboratory. The SUV is outfitted with equipment to record global positioning, vehicle speed, and vehicle acceleration. The vehicle also is equipped with a state-of-the-art eye-tracking system that consists of two infrared light sources and three cameras mounted on the dashboard facing the driver. These cameras and lights are small and not attached to the driver in any manner. The cameras are synchronized to the light sources and help track the head position and gaze of the driver.
An eye-tracking device in the FRV helps researchers study where drivers are looking when they drive through various roadway environments. There are three additional cameras mounted on the exterior of the vehicle’s roof, directly above the driver’s position, for capturing the forward driving scene. The cameras capture the panoramic view of the driving scene in front of the vehicle, and provide a forward view that is 80 degrees wide and 40 degrees high. The forward view area reaches from the left side of the windshield to a portion of the right side.
A second FRV, a 2011 sedan, was acquired by the Human Factors Team on February 2015. Similar to the SUV, the sedan is instrumented with equipment such as a state-of-the-art eye-tracking system. The eye-tracking system installed in this vehicle is comprised of three infrared cameras mounted on the dashboard and a forward-scene camera (mounted to the right of the rearview mirror) to record the visual scene as viewed by the driver.
Using these instrumented tools, researchers are able to record and analyze multiple vehicle measurements, such as steering wheel angle, vehicle speed, accelerator position, brake usage, distance traveled, use of turn signals, use of steering wheel buttons, and other variables. Collecting these measurements from roadway experiments allows researchers to study and better understand driver behavior and performance.
The Sign Design and Research Facility is another facility at the Human Factors Laboratory. It’s commonly referred to as the "sign lab." This facility consists of a 60-inch (152-centimeter) light-emitting diode/liquid crystal display (LED/LCD) high-definition television connected to a computer control center. A new infrastructure design software suite was recently added to the laboratory that will enable the rapid development of interactive static or dynamic roadway simulation environments. This new software will allow for better sign development, replication of existing signs, and the development of realistic roadways (including the use of existing geographic information system (GIS) data).
The sign lab enables researchers to present traffic signs to participants in a controlled environment. When developing new traffic signs, researchers need to determine the maximum distance at which participants can recognize and comprehend signs.
To do this, a participant sits at the computer and looks at the screen as a researcher, sitting at the control panel behind the participant, displays a sign of a small, distant object and then enlarges it using specially designed software so that its appearance approximates the way it would be viewed as a vehicle approaches the sign at a specified speed. The researcher then uses the size of the image at the moment the participant says that he or she recognizes it to approximate the sign's recognition sight distance. The computer precisely controls the sign display duration and image size, and measures the participant's reaction time. The researcher generally records sign comprehension using open-ended questions relating to the participant's understanding of the traffic sign. For example, the research might ask, "If you were driving and saw this sign, what action would you take?"
When developing new traffic signs, researchers need to determine the maximum distance at which motorists can recognize and comprehend a sign. The FHWA sign lab, shown here, enables researchers to present traffic signs to participants in a controlled environment and study their responses.
Recently, FHWA researchers at the sign lab conducted two studies funded by the Traffic Control Device Consortium Pooled Fund Program, which combines States' funds into a pool for Federal research. The first study evaluated identification signs at freeway interchange approaches and the efficacy of the signs at providing motorists with information based on business logos. Currently, the Manual on Uniform Traffic Control Devices (MUTCD) limits the number of business logos on a single interchange approach sign to six. Whether increasing or decreasing this number would produce favorable results was one aspect of the study. The research also evaluated the effectiveness of using businesses' logos versus standard highway sign text.
The researchers showed 103 participants multiple combinations of four-panel, six-panel, and nine-panel signs, an example of which is shown below.
An example of the four-panel, six-panel, and nine-panel signs.
They displayed the signs on the television screen at a simulated distance of 121 feet (37 meters), approximately half the minimum legibility distance. Results suggested that participants were less able to identify specific business logos accurately compared to standard text on highway signs. (See FHWA Traffic Control Device Pooled Fund Program for more information.)
Participants also needed more time to identify artistic logos. Across each of the panels, identification accuracy was higher starting at the top of the sign and shifting downward from left to right. Additionally, more signs on a panel resulted in more eye glances away from the simulated road. Results from this study showed that any benefit of providing drivers with more service information, such as nine-panel signs, is outweighed by the potential risk of increasing driver distraction. The second study performed in the sign lab examined the legibility of multiple alternatives of symbols listed in the MUTCD. The alternatives were either currently used internationally, were State specific, or were generated by the lab or elsewhere. Each research participant evaluated each symbol. The team exposed the participants to scenarios containing each of the sign alternatives for each of the sign groups.
For legibility testing, the researchers used software designed to increase the size of the sign gradually, simulating how the sign would appear as a motorist drives toward it at a specified speed. The researchers then measured the legibility distance for each sign. Following each scenario, the team recorded the participants' comprehension using open-ended and multiple choice questions, and by the participants' rankings of how well they thought the signs would work.
Results showed that some alternatives clearly performed better than others, while other comparisons were not as definitive. For instance, under the multiple choice questions, Alternative 2 of the WEAVE Symbol (DIVERGE) clearly outperformed the three alternatives, garnering correct responses 95 percent of the time. In the case of the four alternatives for the TRUCK ROLLOVER WITH ADVISORY SPEED LIMIT sign, however, the results revealed no statistically significant differences in performance.
In partnership with the National Highway Traffic Safety Administration (NHTSA), the Human Factors Laboratory houses a MiniSim™ driving simulator, a part-task simulator consisting of a quarter-cab setup that includes an adjustable driver's seat; driver controls, such as pedals and a steering wheel; and a meter cluster that includes a speedometer. The MiniSim™ has three 42-inch (107-centimeter) forward-display LCD televisions, software, and computers for generating driving scenes and controlling vehicle dynamics.
The MiniSim™ driving simulator shown here enables researchers to conduct low-cost studies to answer specific questions or preliminary research prior to a larger scale test.
The MiniSim™ is useful for evaluating driver performance in simple environments, such as various infrastructure-related studies that do not require the full immersion of high-fidelity driving simulation. This tool enables researchers to conduct low-cost studies to answer specific questions or to conduct preliminary research prior to a large-scale simulation or onroad research.
A recent study using the MiniSim™ examined driver performance on horizontal curves of rural two-lane roadways. According to the Fatality Analysis Reporting System, a total of 23,740 fatalities resulted from run-off-road crashes on the horizontal curve sections of rural two-lane roadways from 2005 to 2009; an average of 4,748 fatalities per year. An analysis of the National Motor Vehicle Crash Causation Survey suggests that a driver who is familiar with a roadway is twice as likely to be involved in a run-off-road crash as one who is unfamiliar with it. In addition, a driver who is in a hurry is 3.2 times more likely to be involved in a run-off-road crash than a driver who is not in a hurry. Also, an inattentive driver is 3.7 times more likely to be in a crash than an attentive driver.
The research team examined possible procedures for establishing a driver’s familiarity with a roadway, eliciting a state of distraction because of being in a hurry, and determining the effect of these factors on driver performance on rural two-lane horizontal curves, as compared to baseline conditions. Measurements included vehicle speed and lane positioning.
Results indicate that the methodological procedures were effective at simulating the precipitating events and might be useful in future experiments by providing realistic driving situations for the development of dynamic traffic control devices using simulation.
Pedestrian and bicycle rider fatalities account for a significant share of motor vehicle crashes each year. In 2013, nearly three-quarters (73 percent) of these fatalities occurred in urban areas where traffic densities are high and pedestrian and vehicle conflict points are numerous. Vehicle to Pedestrian (V2P) technologies offer the opportunity to reduce pedestrian and bicycle rider fatalities by detecting imminent pedestrian and bicycle vehicle collisions through the application of advanced surveillance and communications.
Virtual reality is defined as a computer-generated environment that gives the user a sense of being in a displayed virtual world through realistic images, high quality sound, and the ability to interact with the virtual world. An important attribute of virtual reality is giving the user a feeling of immersion in the simulated environment. The application of virtual reality technology offers the opportunity to incorporate a broader range of behavior of pedestrians and drivers into roadway experiments without placing these subjects at risk. For example, virtual reality technology could operationalize the linking of a driving simulator with a pedestrian simulator to create a shared virtual environment where vehicle drivers and pedestrians could interact without the risk of a physical collision.
In the VR Lab, two headsets from different manufacturers are being tested. The two devices are very similar with respect to display resolution and field of view. The systems principally differ with respect to controllers and tracking systems. Current work has focused on creating a scenario to support V2P research. Parts of the Turner-Fairbank campus have been modeled along with the new signal installed by the Saxton Operations Research Laboratory.