Predicting the walking speed of pedestrians on stairs

Abstract In this paper, we propose a framework in which the behaviour of a pedestrian is predicted based on the characteristics of both the pedestrian and the facility the pedestrian uses. As an example of its application, we develop a model to predict the walking speed of a pedestrian on stairs. We examine the physiology and biomechanics of walking on stairs, and then develop a model that predicts walking speed based on the weight and leg extensor power of the pedestrian, and the gradient of the stairs. The model was calibrated by experiment and validated by observations. The proposed framework establishes the importance of bridging the two types of characteristics: those of a pedestrian and those of the facility the pedestrian uses. Also, the developed walking speed model is useful for simulating how the design of stairs affects pedestrian circulation.