Using a Virtual Reality Walking Simulator to Investigate Pedestrian Behavior

This highly immersive virtual reality protocol allows us to investigate how people adjust their emotion during crossing. Virtual walking on a treadmill makes it possible to fully captures the reciprocal relationship between perception and action. This message that can provide insight into the field of behavioral ecology as well as allow researchers to explore our passions in pedestrian safety and autonomous vehicle development.

It is helpful to use diagrams to visualize the crossing situation so that the researcher may calibrate the parameters correctly to suit the purpose of his or her study. Demonstrating their procedure will be Hui Li, a graduate student from my laboratory. Begin by setting the parameters on the Walking Simulator using the Config Directory.

In the car section, set the parameters for the first vehicle. Set type to one for sedan, two for bus or zero for no vehicle. Next, set speed in kilometers per hour and distance to the desired value in meters.

Complete the second car section by setting the same parameters. The road section contains parameters for lane selection, set the parameter lane to one to use the lane closer to the pedestrian starting position or two for the lane further away. In the saved section which contains the parameter related to sampling frequency, set the parameter number per second to the desired value in hertz then save the configuration file and exit.

Prepare three practice trial configuration files and create a separate sheet with the list of configurations to be used in the experiment in a randomized order. Recruit participants with normal or corrected to normal vision and ask them to sign a written informed consent form before each experiment. Play an audio recording with verbal instructions of the task to the participant.

And encourage them to ask questions. When ready lead the participants to the treadmill and harness the stabilizing belt to the participant’s waist. Instruct them to hold the hand rails at all times during the experiment.

Ask the participants to practice walking on the treadmill with the belt on while holding the hand rails. Once the participant is able to walk on the treadmill comfortably, double-click the Executable Simulator to begin the Walking Simulator Program. Instruct the participant to wear the headset, providing assistance as needed.

Check for both comfort and stability with respect to head turns. Calibrate the headset so that the black and white cartoon crosswalk is properly aligned with the participants view. Inform the participant that the first practice trial will occur without any vehicles and begin the trial.

Enter the first practice trials configuration number in the text box on the bottom of the screen and click the start button. Instruct the participant to look straight ahead, get ready when they hear ready and begin walking when they hear go, then give the verbal cues ready and go. When finished with the first practice trial perform the second practice trial that introduces vehicles without walking.

For the third practice trial, inform the participant that it will involve two vehicles coming from the left side and that he or she should attempt to cross the road between the two vehicles. Enter the third practice trial number in the text box and click the start button. To perform the Virtual Walking Experiment, type in the first configuration number from the data sheet on the text box and click start.

Perform the simulation in the same way as the final practice trial and record the results next to the configuration number on the experiment sheet. The Walking Simulator was used to determine if the initial distance from the curve to the intersection point affects the approach velocity of participants. The velocity of young adults increased throughout the approach however, when the initial distance was short, they slowed down at the beginning of the trial and sped up continuously.

For children, vehicle size affected the velocity profiles and crossing position induced by the initial distance. Post-walk analysis showed that children sped up throughout the approach however, when they crossed between cars they slowed down at the beginning of the approach for the near initial distance. When children crossed between the buses, their speed neither increased nor decreased at the beginning of the approach for the near initial distance.

It was found that children’s times of intercept increased significantly as the initial distance increased from near to far however, when crossing between buses, children’s times of interception were not significantly different between near and intermediate initial distances. In order for this method to provide accurate information about crossing behavior, the visual field scenes for the virtual reality headset must be properly calibrated. This technique allowed researchers to develop mathematical models of pedestrian behavior which were used to study concepts such as bearing angle and affordance.