Chenshu Wu, PhD | The University of Hong Kong;
Competition Sponsor: Research Grants Council of the Hong Kong Special Administrative Region, China
Awardee year: 2022
Falls are the leading cause of accidental injury death worldwide for older adults. Globally, about 70 thousand fatal falls and over 37 million severe falls requiring medical attention each year, resulting in substantial financial costs (totaling over US$50 billion in the US alone). The problem is aggravated by the fact that millions of seniors live, who are most vulnerable to falls, which can be undetected and helpless for hours or days. Existing solutions to fall detection require seniors to wear a pendant or other sensors. Wearables, however, do not work well for the elderly, who typically find it cumbersome or forgettable to wear them. More attractive wearable-free solutions mostly require training and are either inaccurate or invasive today. This proposal seeks to transform device-free passive fall monitoring solutions with two goals: to detect falls accurately and responsively without prior training, and to monitor motions and vital signs before and after a fall, both in a non-contact and affordable manner. We propose a smart sensing system capable of continuously monitoring an older adult’s motion, breathing rates, and walking speed, which allow reliable fall detection and offer unparalleled insight of fall-related activities. We achieve so by empowering the prevalent and accessible smart speakers with non-invasive acoustic sensing and learning. With improved user comfort and decreased costs, more people worldwide will benefit from contactless fall monitoring. Importantly, the accumulated fall-related data can provide physicians the medical insight to avoid falls. Ubiquitous fall monitoring today can prevent fall tragedies and improve healthy aging tomorrow.