Objective: Over the past decade, the research community has made major advances in technologies for detecting risk of falling in elderly, which led to assistive systems that can help to improve physical strength and limit isolation. However, it is still expensive to extend these systems to a multi-user performing a clinical test, such as the Timed Up and Go (TUG). In this paper, we therefore propose a new avenue, a radar-based system that transcends these limitations.
Results: The proposed methodology is evaluated using data from real-world scenarios. It is found that distinguishing two participants performing a TUG test at the same time is more accurate when they are walking in opposite direction compared to perfectly synchronized walking in the same direction. However, although it was more challenging to distinguish two participants in this latter scenario, our results show that when they are offset by a certain distance, the proposed algorithm can track them separately. Our goal was therefore to relieve congestion in medical settings by investigating the feasibility of administering a clinical test with more than one participant at the same time thanks to an UWB radar sensor.