Reliable Human Fall Prediction Utilizing Robot Fall Prediction Algorithm

The increasingly aging population has inspired a great deal of research in the prevention of fall injuries. A pragmatic and accurate technique to predict fall incidence, along with a corresponding mobile phone app, is proposed in this paper. The technique aims to integrate the benefits of traditional medical history based paradigm and non-historical paradigm and overcome the disadvantage on both paradigms, i.e. too long alert time of the history based paradigm and too short alert time for the non-historical approach. The breakthrough comes from the application of robot fall prediction algorithm to the human subjects. In our previous paper, a similar mobile app analyzes single leg motion to predict the fall of the carrying individual with a little too short alert time. Hence our previous method falls more on the side of non-historical paradigm, not a real hybrid or midway technique. The technique given in this paper represents the real solution, though the human side data is not presented. The missing data will be provided in our next paper.