Using AI To Make Train Boarding Smoother

DZone 's Guide to

Using AI To Make Train Boarding Smoother

Explore an AI use case that automates the train boarding process using LED strips on the platform.

· AI Zone ·
Free Resource

Trains are an increasingly popular mode of transport, but frustrations abound about the lack of punctuality of the network. Indeed, data suggests that up to 70% of British trains are late during rush hour. A team of researchers from the University of Lancaster believes that the problem is exacerbated by the slow boarding process.

They argue that the crush as new passengers try to get on at the same time as existing passengers try to depart can contribute to 50% of the delays experienced on the train network. As such, they have worked to improve this 'platform-train interface' to make the whole process smoother and more efficient.

The researchers believe that AI can help and have developed a system that captures the images taken by the CCTV cameras attached to each carriage and on the platforms themselves to detect not only the number of passengers but the type (i.e. those with pushchairs or bicycles). It's then capable of measuring both the movement and position of these in relation to the train doors, and it uses this data to predict any potential problems with boarding.

Smoothing the Flow

The system is able to detect how busy each door is as people get ready to depart and board. If it believes that a particular door is going to be incredibly crowded, it warns those waiting to board and recommends moving to another door that is less busy. They are hoping to automate this process using LED strips on the platform that change color depending on the state of each carriage door.

While it's not something that's in use yet, such usage of LED strips to inform passengers is something I've covered before. The system was developed by the Edenspiekermann design agency, who developed an LED display for each platform that not only provides the usual timetable information but can also provide passengers with details such as the congestion on the train itself.

In addition to providing passengers with information on the congestion of the train, it also provides them with information on where they should stand on the platform so that they are in the right spot for specific carriages, whether it's first class or disabled (or indeed the bicycle) carriages.

The display itself is 180 meters long and spans the length of the platform. Numbers within the strip highlight whether the carriage is first class or standard, and there is also a marker highlighting the exact position of the doors. Symbols will highlight things such as the disabled and bike carriages, but also information such as carriages fit for buggies, large luggage, and even the quiet carriages.

The capacity of the train will be detected using infrared sensors within each train that will be able to gauge how full each carriage is. If the LED strip shows as green, then seats are available. A yellow display reveals that the carriage is pretty full, whereas a red display suggests you're out of luck.

Despite the system winning design awards, however, it doesn't appear to have been rolled out across the Dutch rail network, suggesting that even the most useful innovations face countless barriers to adoption.

ai and trains ,ai in everyday life ,ai use case ,artificial intelligence ,led lights on train platforms

Published at DZone with permission of Adi Gaskell , DZone MVB. See the original article here.

Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}