Current Interests

This study uses small, portable inertial measurement units to measure human motion.

Currently, accurate measurements of human kinematic motion data, particularly for data to be used for clinical purposes, are generally only available using lab based equipment, typically a camera and marker setup. These systems can be very accurate, but they have a limited workspace and can only be used to measure laboratory based movement tasks. Some measurements are difficult to make in laboratory situations. When measuring human walking, for example, the subject may be told to walk at a comfortable speed. This so-called self-selected gait has debatable relevance to the subject's lifestyle, which can be measured by self reported "activity levels". It would be ideal to instrument a person with non-obtrusive sensors to be worn all day, collecting kinematic data on the subject which can be later analyzed for metrics of activity levels (walking speed, total number of steps per day). More importantly, the data can yield important clinical measurements such as step width and step width variability of a subject, which are often used to quantify a person's risk of falling.

This study looks at using inertial measurement units to measure the kinematics of human motion. Inertial measurement units require proper filtering techniques to yield usable data. We plan to exploit knowledge of the dynamics of walking using a kalman filter to perform model based state estimation.