Introduction: Artificial Intelligence and Subway Systems
Artificial intelligence and subway systems: For decades, subway systems kept the world’s major cities ticking over. Who could imagine New Your City without the subway or London without the tube? Cities like Singapore and Shanghai would likely come to a standstill without their highly efficient subways systems.
The coronavirus pandemic put a temporary halt on public transport, including subway systems. As citizens were asked to stay at home and companies introduced remote working, these arteries underneath our cities became quiet. As the world opens up again, subway systems are once again becoming critical in cities around the world.
However, this time, the trains run differently. Artificial intelligence (AI) is changing every aspect of modern subway systems.
AI and the Ticketing Experience
Most commuters are used to paying for their subway far in bulk. Topping up the credit available on a metro pass or an oyster card has become normal. Whilst that system is convenient when you are traveling outside of rush hours, it still leads to queues when things get busy.
Artificial intelligence technology like facial recognition or biometrics including fingerprints can make for a smoother ticketing experience. Moreover, these technologies can also minimize how often someone needs to come in contact with high-touch surfaces.
Shanghai Metro introduced an AI-led ticketing system using voice activation in 2017 and continues to improve its functionality.
AI and Subway Route Planning
To the average traveler, it may seem like subway trains only have one track to travel along a given route. However, in many cities, trains have alternatives. Whilst there may be a preferred track which they would usually use, there may be others connecting the same two stations.
In case of a delay on one route, because of a breakdown or particularly crowded platforms, AI-driven by data gathered in real-time can suggest another route. The result? The train stays on schedule and avoids becoming stuck behind other trains.
Anyone who has lived in a busy city knows this scenario: you are speed-walking through a subway station as fast as you can to try and make your connection. Just as you are rushing down the stairs to the platform, the doors close and the train pulls away.
With the help of AI, you would have made your connection. The train you are connecting to would receive a notification that your first train has just pulled in. Based on realistic data of how long it takes to walk from one platform to the other, the second train knows how long to wait.
You benefit from no longer having to run through subway stations, pushing people out of the way. Your commute becomes smoother.
AI and Crowd Management
Here is another all-too-familiar scenario for regular subway users: it is rush hour and trains and platforms are packed with travelers.
When a train arrives, passengers struggle to alight. More passengers push in, and then the doors will no longer close because someone’s jacket sleeve got stuck. Cue opening and closing of doors, leading to wasted time.
With the help of technology, managing the crowds in subway systems becomes easier. Predictive AI can have an accurate grasp on how many people will get off a train and will only allow the number of passengers on the platform that will fit into the train car.
During the pandemic, New York City’s transit officials tested another use for AI and subway crowd management to check for compliance with mask mandates. Applications like this may have implications for privacy regulations, but they also have great potential.
AI and Better Train Cars
Cities like Singapore already offer different types of train cars on their subways. The trains traveling to Changi airport have cars with limited seating but additional space for luggage, for example.
AI can take this concept further. Understanding how passengers use the cars, how many seats are required, and how much standing space allows managers to put the most suitable cars on a line.
AI and Predictive Maintenance
Predictive maintenance is no longer a pipe dream for Hong Kong’s subway engineers. Operator MTR has been using an algorithm to schedule nightly engineering works and maintenance crews since 2014.
The algorithm has a view of the entire system that would normally require experts from up to six different areas to come together. But it does not work without human input. When the software was developed, programmers consulted the experts to find out what guided their decision. They based their code on these expert insights. As a result, the algorithm acquired the equivalent of years of human knowledge within days. Saying that its suggestions are still reviewed by humans before engineers set off for the night.
Conclusion: Artificial Intelligence and Subway Systems
Subway systems are critical to effective transportation in the world’s major cities. AI technology can help the trains run on time by optimizing schedules, routes, and connections. Algorithms help predict and schedule maintenance, creating a better experience for travelers around the world.