Assessment of Human Gait Speed and Energy Expenditure Using a Single Triaxial Accelerometer

Assessing people's everyday life physical activity by accelerometry has gained a large role in the medical, sports and psychology science research in recent years. Walking is the most common everyday life activity and walking speed can be used as an indicator for human physical activity profile. Energy expenditure is on the other hand often used as a dimension while assessing human activity and is strongly related to people's walking speed. We propose an algorithm, which predicts the gait speed based on body accelerations measured on the right side hip by a tri-axial accelerometer. Based on the gait speed the activity is classified and the energy expenditure is estimated. After applying the algorithm to free living track data (both indoor and outdoor), the results show a good agreement between actual and estimated gait speed and energy expenditure.