Current research has shown that miniaturized body worn acceleration sensors can produce comparable results to existing validated clinical gait and balance scales but they are still used mainly in a research setting. The current study shows that it is technically and logistically possible to introduce standardized accelerometry into the clinical practice in the context of an European multi-center setting. The results obtained from the acceleration signals confirm the expectations and provide more information compared to the traditional methods used to assess postural stability. Finally, the correlation between gait speed and age adds to the construct validity of our algorithms. We conclude that the actibelt® platform is a promising technology to be further developed towards a validated personalized fall risk assessment tool set.

Though Center of Gravity (CoG) defuzzification is a well-known and long-standing method in the history of fuzzy systems, because of its computational complexity, its use in the field of modeling of fuzzy controllers is almost nil. From literature, it appears that modeling of fuzzy Proportional Integral Derivative (FPID) controllers is rarely attempted using CoG defuzzification. In fact, none of the FPID controller models are obtained using both two-dimensional input space and CoG defuzzification. The available mathematical models of fuzzy Proportional Integral (FPI) and fuzzy Proportional Derivative (FPD) controllers using two-dimensional input space and CoG defuzzification were due to Arun and Mohan (2017). In this paper, the authors make an attempt to model and design an FPID controller using two-dimensional input space and CoG defuzzification. The incremental control effort produced by the newly developed FPID controller is found by combining the individual control efforts produced by incremental FPI and incremental FPD controllers. The incremental FPI and incremental FPD controller structures are unveiled using two-dimensional input space, CoG defuzzification, Min t-norm, Max t-conorm, and Larsen Product (LP) inference. The applicability and usefulness of the newly obtained FPID controller are depicted with simulation and real-time experimentation.