Hong Kong’s rail giant has implemented two AI-powered systems to help formulate train deployment and crowd control plans for major events at Kai Tak Sports Park, with plans to consider expanding their applications.
The MTR Corporation said on Friday that the two artificial intelligence (AI) tools – a ridership prediction model and intelligent crowd diversion system – had helped disperse passengers efficiently from major events at the park.
The ridership prediction model, developed by the firm and Hong Kong University of Science and Technology, relies on billions of data sets collected from government surveys and MTR operations to predict passenger distribution across the entire rail network.
It also analysed the rail firm’s historical passenger data spanning more than 100 days when concerts and sports events had taken place at Hong Kong Stadium and the Hong Kong Coliseum to generate virtual scenarios after such events.
“We can therefore predict the number of passengers, their travel directions, and the stations and lines they use after an event ends,” said Chan Hing-keung, the MTR Corp’s chief of operations for engineering service and innovations.
“With the prediction, we can cross-check with the analysis by the operations team, and determine whether the frequency of trains could disperse the crowds.”
Launched in July 2024, the model was first used at the end of that month, when the section between Ho Man Tin and Prince Edward on the Kwun Tong line was suspended as the rail operator upgraded its facilities.
