Acoustic Situation Awareness and Its Effects on Pedestrian Safety within a Virtual Environment

The use of personal listening devices is reaching new zeniths among the age group of 18-35 years of age, especially among the student population. Personal listening devices are being used on a daily basis by students to listen to music, podcasts, and audiobooks among many other things. Most prefer to use PLDs while performing other activities such as walking on campus, working out, or while using public transportations. However, the use of PLDs in situations that require focused attention, such as crossing a street, can lead to the preclusion of situation awareness. This research is focused on the effect of PLDs on acoustic situation awareness.

Our first experiment, an observational study, takes place across multiple crosswalks on the Blacksburg campus. This is to identify different crossing scenarios. For example, we selected a crosswalk on Alumni Mall to observe the effect of parked buses; we observed the crosswalk on Main Street and College Avenue to include a signalized crosswalk. Observations were made to identify potentially risky pedestrian behaviors, especially for those using PLDs. Risky behaviors included, not watching both sides of the street before crossing, starting to cross too close to an approaching or passing vehicle, using a mobile device (phone) while crossing, engaging in a conversation with a group while crossing, or making any gestures (behavior of interest). We also made note of the kind of PLDs being used; whether it was an air conduction earphone or bone conduction earphone. We also distributed a survey to pedestrians exhibiting risky behaviors to understand their decisions better, and also understand how often they use PLDs. Lastly, a couple of focus groups are also planned to gain further qualitative data on street crossing behaviors, frequency of use of PLDs, music and listening level preferences, and gesture choices. Observations were recorded using a GoPro mounted on a tripod, for 2-3 hours at the different sites at times of peak foot traffic. A sound level meter was used to record ambient and vehicular dB levels at different points of the crosswalk, and a speed gun was used to measure the average speed of vehicles.

Based on the data collected from experiment one, we plan to conduct a second experiment where we simulate the conditions observed in the real world inside a virtual environment. Unreal Engine is being used to build the environment, populated with vehicles and pedestrians, with changing time of day, and real recordings of spatial audio played through speakers on all four walls of an anechoic room, all with a one: one mapping with the real world. The aim of the second experiment is to provide evidence for loss of situation awareness because of the use of PLDs. A control group would perform street crossings in the virtual environment without the use of any PLDs for a variety of scenarios, and a second group would perform the same actions but while using a PLD (air conduction vs bone conduction) at different listening levels.