Application Title

City:

Hong Kong, China SAR

Organization:

Mass Transit Railway (MTR)

Project Start Date:

1 September 2014

Project End Date:

1 June 2014

Reference:

A. Halvorsen, H.N. Koutsopoulos, S. Lau, T. Au, and J. Zhao. “Reducing Subway Crowding: Analysis of an Off-Peak Discount Experiment in Hong Kong,” Transportation Research Record: Journal of the Transportation Research Board, No. 2544, pp.38-46, 2016. 

Problem:

Several MTR segments exceed loading standards in the AM peak. The MTR wanted to implement the Early Bird policy that offers discounts for trips taken before the AM peak. They wanted to know what type of users represent most of the AM peak traffic and whether these users will be receptive to the new policy. 

Technical Solution:

K-Means was used to cluster users, with 6 groups being identified. For each type of user, the percentage of AM peak and pre-peak trips they represented was calculated before and after the Early Bird policy was implemented.

Datasets Used:

• Dataset 1: Octopus Card trip data, MTR, September and October 2013. 

Outcome:

The Early Bird policy resulted in shifting 2.5% of trips out of the AM peak hour. The policy was only effective in shifting trips from the peak hour to the pre-peak hour for two out of the six user groups, but these two groups represented 85% of all AM trips. 

Issues that arose:

Hong Kong’s Umbrella Movement occurred between the end of September 2014 and December 2014. These protests disrupted bus service, causing an increase in ridership on the MTR. This increase in ridership made it difficult to measure the impact of the Early Bird promotion, meaning that only one month of reliable data could be collected. 

Status:

Operational.

Entered by:

Lauren Wilcox, lauren.wilcox@mail.utoronto.ca