Carrying-Mode-Free Stair Ascent and Descent Estimation using Smartphones

This paper proposes a smartphone-based pedestrian dead reckoning to track stair ascent and descent. By using accelerometer and barometer information, pedestrians’ gait patterns are collected under different carrying modes to detect the stair ascent and descent motions. Also, weather influences are considered to eliminate the barometer measurement errors. We evaluate the proposed method in a seven-floor environment, and we compare our method with a conventional method without multiple carrying modes and weather considerations. The minimal error rate of the proposed method is only $4.2 \%$ when ascending or descending seven floors, and at most $14.4 \%$ error are reduced compared with the conventional method.