Using Zigbee Sensors for Ambient Measurement of Human Gait – Analytical Considerations
Ashi Agarwal, Bruce Wallace, Laura Ault, Julien Larivière-Chartier, Frank D. Knoefel, Rafik A Goubran, Jeffrey Alan Kaye, Zachary Beattie, Neil William Douglas Thomas · 2021
With the aging of the population in Canada and elsewhere, applications of Smart Homes for well-being sensing are increasingly being considered in health care. Many of these smart home networks rely on the Zigbee wireless protocol to connect sensors used to measure various health outcomes. This paper provides preliminary results of gait estimation performed on 3 different residences over 11 months using Zigbee connected motion sensors, with a focus on understanding accuracy limitations induced by the Zigbee communication protocol. The accuracy limitations were also observed in the results from a controlled experiment done with 2 different sets of Zigbee motion sensors. This paper provides an in-depth analysis on root cause of variance in gait estimation at the same time laying out conservative variance estimations caused by different scenarios. The accuracy considerations highlighted by the paper are also applicable for all other time sensitive measures. Results of this paper necessitate further analysis of the use of Zigbee operated sensor networks in the evaluation of time sensitive measures.