Effect of Walking on Escalator to Safety


Student thesis: Doctoral Thesis

View graph of relations


  • Wing Hong William LAM


Awarding Institution
Award date14 Aug 2015


Escalators are widely used in public areas such as public transport and shopping centres. Riding on an escalator is part of the daily life of most people living in urban areas and the ridership is in the order of tens of millions per day for a city like Hong Kong. The accident frequency is the highest amongst all types of accidents in metro system and many of them are associated with the passenger behaviours such as lost balance when walking and carrying large luggage on escalator. Although walking on an escalator is generally perceived to be unsafe, there are different practices related to walking on escalator. Many cities have the practice of commuters standing on one side allowing the passage for walking on the other side. Commuters stand on the left in Hong Kong, London, Washington and Tokyo, and commuters stand on the right in Singapore, Australia, and Osaka. On the Montreal Metro, walking is prohibited but the commuters generally do not follow this rule.
This research aims to find out whether walking on escalator would lead to higher probability of accident. In this research, the key parameter of number of passing between walking commuters and standing commuters on escalator is identified as the parameter representing the risk of escalator accident induced by walking. The data collection, method development and the process of analysis are designed to evaluate this parameter which is compared to the accident record for demonstration of the effect of walking on escalator to happening of accident.
The record of 5,682 escalator accidents in 38 stations over a period of 11 years was studied. In exploring the key factor associated with walking on escalator and its association with accident, connection of various data sets was examined in order to find out the relationship amongst various factors or parameters. The station patronage data captured by the entry and exit gates is used for analysing the crowd density at the time of accident. Field surveys at certain locations in stations were conducted to find out the route choice of commuters for providing additional information as input data in the commuter flow simulation. The data from various sources are connected through commuter flow simulation.
Analysis is conducted at two levels. At station level, statistical analysis is conducted to study the conditions in terms of crowd density when the escalator accidents occur. At facility or escalator level, commuter flow simulation is conducted to evaluate the identified parameters, which are the escalator flow rate and the number of passing between walking commuters and standing commuters. Statistical analysis is conducted on the results of commuter flow simulation to examine the correlation between the parameter associated with walking and the record of escalator accidents.
Processes are organised to collect data, starting with the available record of escalator accidents and the station patronage record captured in the entry and exit gates. The station patronage data, accident records, and the field survey data are analysed to establish the statistical overview and input for flow simulation. In selecting the simulation software, commercial packages are examined as they are the software commonly used in the industry. Commercial packages however are designed for evaluating the time of travelling from one point to another, and not suitable for simulating commuter walking on escalator. Simulation software based on grid-system Cellular Automata is thus developed for this purpose. The parameters associated with commuter movement on escalator are counted using the software. Analysis on these parameters and the escalator accident record is conducted to establish the correlation between walking and escalator accident.
Based on the demonstrated correlation between walking and escalator accident, prediction model can be developed for evaluation of design options for different building layout and provision of escalators. Improvement direction of escalator design is also recommended based on the result of this research.