Using Robot and Electric Drive in Fall Prediction
Wei Ding, William Engel · 2018
The global aging phenomenon has motivated active research in human fall injuries. The fall prevention has hence become a popular topic in health informatics. An effective fall prevention paradigm could save millions of people from injury and avoid considerable casualties. Through comparison studies, detail-oriented simulations, and pragmatic field tests, an effective fall prediction method has been developed by authors. The finding is presented in this paper. Three techniques for fall prediction are discussed in this paper. A comparison technique to mimic the traditional stateless fall prediction techniques, along with an algorithm using artificial neural network, was first implemented in authors' previous paper. Then a robotic scheme was developed to simulate human fall by transplanting a proven fall prediction paradigm for humanoid robots with controlled electric drive systems to human subjects. Due to its simulation nature far from the human fall scenarios in reality, the robotic paradigm has obvious limits in real world applications. It was also used more like a reference framework for our last scheme. Eventually we built the third approach that eliminated the above limitation. The third approach is elaborated in this paper. Our test and simulation have proved its pragmatic superiority over other two approaches, along with vast majority of traditional paradigms.