Challenges in the railway industry

 

The railway industry is a basic industry that transports large numbers of people over medium to long distances through large-scale infrastructure. The most important factors in the railway industry are the safety of large numbers of passengers and the efficient maintenance of large-scale railway infrastructure. The following are the current challenges/issues faced by railway operators in various countries.

 

• The railway industry consists of train drivers, crews, controllers, station staff, track maintenance workers, train inspectors/maintenance engineers, etc. and railway infrastructure consists of trains, stations, tracks, depots, control centres, etc.

• Currently, in most countries, the railway sector, which is a 3D industry, is experiencing a decrease in the number of new people entering the railway infrastructure maintenance field, and the existing workforce is aging. In addition, not only the workforce but also the existing railway infrastructure is aging, and there is a need for advanced/automated maintenance to ensure transport safety.

• In the railway industry, life safety threats include passenger falls at stations, passenger incidents/accidents in cabins, accidents involving people/vehicles at railway crossings, collisions between trains, safety incidents involving track maintenance workers, and the introduction of abnormal objects on the tracks (people, animals, fallen trees, debris, etc.), which can lead to major accidents due to the nature of the railway industry.

• Although railway operators implement various safety measures, the problem is that they lack an efficient response system/technology for safety incidents. For example, even now, even if there are multiple surveillance cameras at each railway station, there are cases where the situation is not noticed because the situation is monitored by human eyes. In addition, there are problems with installing CCTVs along the entire track (due to the short effective range of CCTVs of tens of metres, it is not possible to install a large number of CCTVs along tens to hundreds of kilometres of track, and the huge cost of fibre optics to connect a large number of CCTVs to a remote monitoring room), and communication network issues (supporting large uplinks and handovers) to stream high-definition CCTV images in the cabin to monitor cabin safety accidents at a remote monitoring center.

• In the area of railway infrastructure maintenance, for example, in the case of the Keikyu Express Railway in Japan, maintenance workers patrol the tracks on foot in groups of 2-4 people once or twice a month to visually inspect the tracks for abnormal conditions. This means that real-time track conditions cannot be grasped. In the case of train inspection, train bogie cracks and brake wear are visually inspected by groups of 5 people at the train inspection station every 6 days. Similarly, there is no real-time train safety inspection. The problem is that human inspection is incomplete and the cost of labour for constant inspection and testing leads to long inspection and testing intervals, which threatens the safety of rail transport.

• To solve these problems, the model of installing monitoring cameras at checkpoints, tracks, and cabins and connecting the control/control centre with a Wi-Fi network was initially applied, but issues such as Wi-Fi connection instability, narrow coverage, and providing large and stable uplinks have limited the solution.

• In addition, even if the 5G private network service based on network slicing of mobile operators is used, the quality of railway application traffic may be degraded by B2C traffic of mobile operators' customers, and conversely, if the quality of the sliced 5G network is guaranteed by providing priority to the railway private network, the quality degradation of B2C users is self-evident.

• With the availability of private 5G spectrum, cases of private 5G networks in the railway sector are emerging, which will make a significant contribution to the safety and maintenance of the existing railway sector.

​

Private 5G for Railways

 

Since 2019, governments in Germany and Japan, among others, have begun to open up 5G spectrum previously reserved for mobile operators to industry to enable digital transformation in the industrial sector. In South Korea, the Ministry of Science and ICT began offering private 5G spectrum to the market in October 2021.

 

 5G privides downlink speeds of hundreds of Mbps to several Gbps and uplink speeds of tens to hundreds of Mbps. It also provides ultra-low latency wireless connectivity of a few milliseconds.

 

Private 5G reduces base station investment costs due to the wider coverage of base stations compared to Wi-Fi networks used in traditional enterprise private networks. In addition, Wi-Fi networks use unlicensed frequencies, resulting in poor communication quality in corporate workplaces such as railway stations and airports where there is a lot of public Wi-Fi access.

 

Private 5G networks use private 5G frequencies, allowing the company to use the frequencies exclusively, regardless of Wi-Fi traffic or public 5G traffic. Wi-Fi networks do not support handover, whereas Private 5G natively supports handover with 3GPP technology, which is essential for environments such as trains travelling long distances at high speeds.

The introduction of Private 5G to the railway industry will provide the high-capacity wireless uplinks needed for safety and maintenance applications on the railway.

 

Cameras and 5G UEs can be deployed at stations, tracks, rail crossings, train cabins, drones, and more, and large amounts of camera footage can be streamed in real time over the private 5G network to an AI video analytics server in the control centre, where the AI server can automatically detect safety-related emergencies such as passengers, rail maintenance workers, and pedestrians in the vicinity, and notify the train driver or controller in real time for early response.

 

This enables early detection and early response to life-threatening safety incidents in the railway industry, greatly improving passenger and worker safety.

 

CCTV cameras can also be connected to tracks and crossings with fibre optic cables, but as the number of cameras increases, the cost of cables reduces scalability. Private 5G provides wireless, high-capacity links and transmission reliability comparable to fibre, enabling cost-effective camera expansion.

 

In addition, by installing cameras and 5G UEs at checkpoints, stations, train fronts, and drones, large-scale images can be transmitted to AI servers over the 5G network, where AI servers can automatically detect abnormalities in trains or tracks, enabling full-time and automated train safety inspections and track inspections. 

 

This will help improve the inspection and maintenance of aging railway infrastructure by using cameras and AI to perform inspections and checks on a full-time basis, rather than the intermittent, fortnightly inspections and checks performed by humans. It will also reduce the workload of track maintainers and train inspectors and save manpower, helping to solve the railway industry's labour shortage problem.

 

Due to the shortage of labour in the railway sector and the need to automate and unmanned railway operations to improve the safety of railway transportation, a high-capacity, low-latency, wireless private communication network is essential, and private 5G has recently gained attention around the world.

 

<Automation of railway safety prevention and maintenance based on cameras + private 5G + AI>

 

The safety of large numbers of passengers and the efficient maintenance of large-scale railway infrastructure are the most important factors in the railway industry. Currently, the railway industry is faced with the aging of railway infrastructure, a decline in the entry of new workers, and the aging of the existing workforce, so automated safety management and maintenance are required to minimise the number of labourers.

 

This requires a high-capacity, low-latency, wireless, private communication network. Currently, Japan, Taiwan, and China are adopting private 5G networks that are independent of public 5G networks built using private 5G frequencies in specific areas as a private network technology for railways.

 

There are two main applications that are currently being implemented using private 5G networks in the railway sector. One is life safety for passengers and railway workers, and the other is railway infrastructure maintenance. 

 

In both cases, high-definition cameras are installed in various parts of the railway infrastructure, such as platforms, crossings, tracks, drones, train fronts, cabins, and train checkpoints, and large-scale camera images are uploaded to the AI image recognition server in the control centre or cloud via the private 5G network, where the AI server automatically detects emergencies/abnormal situations on site and notifies the relevant personnel (train drivers, controllers, station staff, maintenance workers, etc.).

 

The main use of the private 5G network is to deliver high-definition video to the AI server through a large capacity uplink.

 

In order to implement a safety management and maintenance automation system in the railway field, a large uplink for high-quality video transmission is essential, but optical cables have limited laying sections and are basically wired, so there are problems with scalability; Wi-Fi has unstable connections and no mobility support; and public 5G networks have a large downlink capacity, which limits the available uplink capacity; however, private 5G networks are optimal for private networks in the railway field because they are 3GPP 5G technology, so the connection is stable and mobile, and the ratio of uplink and downlink capacity can be adjusted by the railway operator.

 

By introducing monitoring cameras, private 5G, and AI to railway infrastructure, railway tasks that were previously performed manually by humans can be unmanned/automated, reducing the number of personnel required for safety management and infrastructure maintenance, enabling operation with a minimum number of elite personnel (compact management), solving the labour shortage problem in the railway industry and enabling efficient management.

 

In addition, by replacing track inspections and train inspections, which were previously performed manually, visually, or on foot by humans, with cameras, private 5G networks, and AI, precise and constant maintenance (predictive maintenance) of railway infrastructure (tracks and trains) is possible, which maximises the reliability of railway infrastructure.