‘Twalking’ is a slang term that refers to someone who is walking while they text or tweet, with all their attention focused on their mobile device rather than the world around them. As a result of this new phenomenon, the rate of death and serious injury in the US caused by twalking has tripled in the last seven years.
In a bid to prevent further fatal accidents, researchers from the Data Science Institute at Columbia University are designing an intelligent headphone system that warns pedestrians of imminent dangers.
— The Data Science Institute at Columbia University (@DSI_Columbia) December 11, 2019
The headphones make use of intelligent low-power signal processors and an array of miniature microphones to detect the distinct sound of approaching vehicles. The user’s smartphone also works in conjunction with the headset to utilise machine-learning models that classify hundreds of acoustical cues from city streets and nearby vehicles. Once the sound of an on-coming car had been identified, the headphones send the user a warning alert.
In 2017, the project received a grant of over R14-million rand ($1.2-Million) by the National Science Foundation and since then the team has received several honors including the best presentation award from the IEEE conference and the best demo award from the ACM conference.
“We are exploring a new area in developing an inexpensive and low-power technology that creates an audio-alert mechanism for pedestrians,” says Fred Jiang, a Data Science Institute member and an assistant professor of electrical engineering at Columbia Engineering.
The headphones are currently being tested and refined in the busy city of New York, with the team of researchers hoping to create a prototype that can easily be transferred to commercial use.
Feature image: The Data Science Institute at Columbia University