Electric scooters (or e-scooters) are a fairly common sight these days in most big cities around the world. While popular among some for making it easier to get around the city quickly, many also consider them to be a huge nuisance. I remember when they appeared seemingly out of nowhere in Columbus, Ohio, popping up everywhere like mushrooms. Suddenly the sidewalks were littered with e-scooters. And they were annoying – to step over, to avoid as they raced down the sidewalk. I’m sure drivers didn’t like them either – weaving in and out of traffic, darting across the road.
But what makes them particularly dangerous to pedestrians is how quiet they are. You can’t hear them coming, and they can go quite fast. They regularly make me jump as I walk around the city, rushing suddenly past me on the sidewalk.
Electric vehicles faced a similar safety issue when they started to become mainstream. Without the classic motor sound, they were dangerously quiet. And so manufacturers added all sorts of noises to them, from artificial engine sounds to more sci-fi sounding hums. These continuous external sounds are called “acoustic vehicle alerting systems” or AVAS. Should AVAS be added to e-scooters as well?
Researchers used virtual reality (VR) to test whether adding AVAS to e-scooters might make them easier for pedestrians to detect (Walton et al., 2025). In VR, 63 participants attempted to detect the presence of e-scooters in three virtual locations: a city park, an open urban space (concourse), and a busy city road. They completed three tasks in these environments. For the first, participants were asked to press a trigger as soon as they heard an e-scooter approaching from behind them. For the second, participants listened for e-scooters approaching from behind and indicated when they heard the e-scooter start to slow down, and then when they heard it stop. And finally for the third, while two other e-scooters circled in front of them, participants again indicated when they heard a third e-scooter approach from behind.
The study tested three AVAS sound designs (continuous, impulsive, and mixed) at two volumes (56 dB and 66 dB, measured at 2 m). The continuous sound is similar to a held chord, the impulsive one contains repetitive bursts of sound, and the mixed one contains both the held and pulsing sounds. For the second and third task, they also examined whether making the AVAS reflect the speed of the scooter (faster resulting in louder and higher-pitched, slower resulting in quieter and lower-pitched) could help make speed changes easier to detect.
Concerning the first task, when the e-scooters had no added sound, participants failed to detect approaching e-scooters 90–97% of the time, indicating that the e-scooters were virtually inaudible. When an AVAS was added at 66 dB, missed detection rates fell to 3% or less (except for the mixed AVAS on the busy city road, which was 18% – still a huge improvement). The 56 dB AVAS was more or less effective depending on the noisiness of the environment: average missed detection rates were just 9% on the quiet park path, 54% in the noisier open urban space, and 85% on the loud, busy city street. Still better than no added sound at all, but nowhere near as effective as 66 dB.
For the second task, participants had a much easier time identifying when an e-scooter had come to a complete stop when the AVAS was speed-dependent (and therefore fell silent at zero speed). In fact, they were able to indicate that they heard the scooter stop over a second faster than when the speed-independent AVAS was used. The speed-dependent AVAS also made it significantly easier for participants to detect approaching e-scooters in the presence of other e-scooters (as tested in the third task).
The type of sound design (continuous, impulsive, mixed) did not have much impact on the results, besides the mixed sound performing slightly worse during the first task (but still quite well at 18% missed detections).
To make e-scooters detectable to pedestrians, and therefore significantly safer, add a speed-dependent acoustic vehicle alerting system (AVAS) of at least 66 dB at 2 m (as opposed to 56 dB at 2 m – the minimum required volume for quiet running vehicles specified in UNECE Regulation 138). By speed-dependent, I mean that the volume and pitch should reflect the speed of the e-scooter: rising as it accelerates, falling as it slows, then shutting off when the scooter stops.
Walton, T., Torija, A. J., Hughes, R. J., & Elliott, A. S. (2025). Evaluation of auditory alerting systems for safe electric scooter operations. Scientific Reports, 15(1), 3424.
road safety, electric scooter, e-scooter, acoustic vehicle alerting systems, AVAS, pedestrian safety, accessibility, urban soundscape, sound design, warning sounds, virtual reality, VR, behavioral study
No posts
Comments
Nothing yet. Say the first thing.
Sign in to join the conversation.