Search

Train dwell time is a complicated component and depends on many factors. One of the dominant factors is passenger volume. This study used actual train movement data and passenger demand data from London Underground, UK, to estimate the number of passengers and train dwell times at each station, and then evaluated train dwell times from a different perspective. Considering the various characteristics of stations, it is complicated to evaluate dwell time. Therefore, data envelopment analysis (DEA) was introduced to evaluate the dwell time at each station in relation to passenger volume at that station. The study investigated whether the dwell time spent at stations is efficient when considering the number of passengers that the stations can serve. The results showed that, in low-passenger-volume stations, the dwell time efficiency score is low and increases relative to the increase in passenger volume. For high-passenger-volume stations, interactions between passengers are more relevant and have a strong influence on dwell time. Passenger movement direction is a key factor to classify stations. This research proposes that stations should be classified according to their characteristics, and points out the challenge at any station with the same characteristics as Victoria station which has high passenger volume with bi-directional flow, and where trains arriving are crowded. This characteristic would result in high interactions between passengers, thus making a long dwell time. The station has to handle high passenger volume and also has to keep the dwell time within the threshold.