51爆料网

Skip to main content

51爆料网 Engineering Expert Partners with Comcast on Fall Detection System

 Moeness Amin, PhD

It鈥檚 been nearly two decades since Moeness Amin, PhD, director of 51爆料网鈥檚 Center for Advanced Communications, began conducting research on sensing through walls using radar. His early work focused on the use of radar imaging technologies to benefit defense, security and the criminal justice system, allowing soldiers and law enforcement officers to map building interiors and locate and apprehend outlaws and adversaries inside enclosed structures. Throughout the past five years, Dr. Amin鈥檚 research has evolved to consider civilian applications for this technology. More specifically, with the support of a number of major grants, Dr. Amin is examining the use of radar to detect falls and monitor the elderly in their individual living spaces. His work has been featured in a variety of national and international news media, from The Atlantic and The Wall Street Journal to the Daily Mail and PBS.

In 2017, Dr. Amin鈥檚 research caught the attention of global telecommunications conglomerate Comcast, which proposed a partnership that would link his work to their smart home technology. Dr. Amin says, 鈥淭he fall detection system could work like a home security system, where there鈥檚 a one-time cost for set-up and then a monthly service fee for monitoring.鈥 The next step in the process will be determining the most reliable algorithm for implementing the fall detection technology. 鈥淲e need a machine- learning algorithm that can detect the difference between a fall and human daily activities, and does not confuse falling with similar motions, like sitting, so that costly false alarms are prevented,鈥 he explains. The ultimate goal of Dr. Amin鈥檚 research in this area of healthcare technology is to make a radar that is personalized and specific to the individual being monitored, which is a form of a 鈥渃ognitive radar.鈥 It would be self-learning and would tune its motion classification parameters as it continues observing the person, delivering superior performance to a radar designed for a generic person.