Communicating to trains is an increasing requirement in the digital age, not just in the UK but all around the world. There are many reasons why this demand has emerged and they span from personal habits of wanting to constantly ‘keep in touch’ right through to the sending of safety critical instructions for ongoing train movement.
Can this broad swathe of need be provided by a single backbone of radio connection? Or do there have to be different technologies for different requirements?
It’s an interesting debate. One firm in particular has given the challenge a considerable amount of thought and has produced an emerging number of technical developments over the past few years, including the roll out of on- train networks and communications systems in many countries.
Wi-Fi on trains
Nomad Digital has been in existence since 2002, its starting point being to engineer a Wi-Fi connection to trains for passenger usage. A number of train operators (TOCs) within the UK signed contracts for this provision and, with the ensuing roll out, the strengths and weaknesses of the architecture began to emerge.
Unlike rival companies that chose to rely on satellite communication (itself having limitations when encountering tunnels and other shadow spots), Nomad elected to use the terrestrial public 3G (and latterly 4G and soon to be available 5G) networks as the communications path. Added to this train-to-shore solution, Nomad equips the train with its own internal Wi-Fi network thus making the train a kind of moving ‘hotspot’.
There are two common ways of linking the Wi-Fi access points along the train: either by providing an internal Wi-Fi system and a series of radio ‘inter carriage links’ or by using the through-train wiring for distribution of the radio signal to each coach. The latter is preferred as the unlicensed Wi-Fi band commonly used for inter-carriage networking is subject to increasing levels of interference and performance can be affected – particularly if it is a long train.
The availability of train network wiring can vary considerably; easy for new trains where the requirement has been built in, less easy for older stock where imaginative ways of multiplexing on to existing train wires might be needed.
A Rail Engineer article on how an on-train Wi-Fi service was achieved and deployed appeared in issue 73 (November 2010). The popularity of the product soon produced its own problems as the available bandwidth could be quickly swamped by users having extended voice conversations and downloading large files.
Such a situation could earn the service a bad name and it was ‘pot luck’ as to how successful a user would be, much being dependent on the level of business on the train and the type of passenger on board. The quality and coverage of the mobile cellular networks used to carry all data traffic to and from the train was another determining factor.
Many of the UK TOCs are now providing train connectivity and, indeed, it is often a franchise requirement. Nomad provides this service for Virgin West Coast, Heathrow Express, CrossCountry, East Midlands Trains, South West Trains, ScotRail and First Great Western. A recent contract signed with Eurostar will equip both the refurbished Alstom fleet and the new Siemens Velaro trains with both Wi-Fi and entertainment services.
Nomad systems have also been deployed in many other countries including Queensland Rail in Australia, Canadian VIA Rail, Amtrak in the USA, LEO Express in the Czech Republic, the PKP intercity network in Poland, NSB in Norway, the NS intercity trains in Holland, and the S-Train network in Copenhagen. All have been duly promoted as part of the Nomad marketing campaign and some include the added provision of passenger information services (PIS) or entertainment content delivery to passengers’ laptops, tablets or smartphones.
The demand for enhanced passenger communication is without doubt a growing business. The challenge is how to provide this within the bandwidth limitations of the external mobile network providers whilst still providing an acceptable quality of service.
The vision of an always-available unlimited Wi-Fi is challenging to deliver and the connectivity has to be managed so as to provide an acceptable level of service shared by all passengers. Offering alternative on-screen services in each coach is expected to reduce demand for unlimited usage from personal mobile devices.
Current thinking
Train connectivity is much more than just providing a connection for travellers’ laptops, tablets and smartphones. It is emerging into an on train ‘service delivery platform’ (SDP) for all communication needs including real time passenger information, staff enterprise applications, news flashes, entertainment and advertising.
Worldwide, only 2-3% of trains are fitted with internet connectivity and Nomad is estimated to have more than 45% of market share. In the UK, roughly a third of trains have on-board Wi-Fi but this is set to grow and is likely to increase to 100% over the next five years. This will mean increasing bandwidth demands despite attempts to moderate passenger usage by providing targeted and relevant on-train information and content. So where will this come from?
The Nomad view is that overall availability of spectrum is likely to change. The days of allocating specific chunks of frequencies to dedicated users are considered to be numbered and therefore the possible successor to GSM-R will need to be designed around a shared radio service. The same will apply to military applications, the emergency services and civil aviation.
In parallel, the public mobile networks will move to IP technology and these will be the backbone for all user groups. Some railway engineers and operators will struggle to get their heads around this but the reality of spectrum value and utilisation is such that having chunks that are lightly or inefficiently used will be unacceptable to the regulation authorities, service providers and governments.
The design and safeguarding of radio services from interference and cyber attack will clearly be a large part of this challenge.
For railways, there is also the problem of signal strength both at the trackside and on trains, and whilst 3G and 4G service providers continue to enhance coverage, it is unlikely that 100% will be available for reliable high-quality rail usage. Companies such as Nomad Digital must therefore be prepared to partner with the public network providers to enhance the base station infrastructure where necessary.
The vexed question of tunnels remains, as these require expensive radiating cable provision to give service. If the transit time at line speed is short, it is unlikely to be financially justifiable but it will depend on usage aspirations for both public and in-house demand.
Operational use
Whilst plying the travelling public with ever-more-comprehensive information by using the connectivity available, why not use this for train borne operational use? Nomad has come up with the acronym ROCM (Remote Online Condition Monitoring) which opens up a host of possibilities for intelligent fleet management and energy savings.
An obvious application is linking a train data bus to a fleet control centre so that data relating to the performance of various on-train systems, including alarm gathering, fuel consumption, speed and braking, that is currently captured for historical purposes, can be diagnosed much earlier thus prompting immediate action or attention at the next depot visit.
The advent of DAS (Driver Advisory Systems) is dependent for success on reliable communication to and from timetable and real time train positioning data, so an enhanced connectivity link will prove invaluable to the performance of such systems.
Countries that have invested in these operational communication links include:
» Portugal, where a joint venture with EMEF (the Portugese rolling stock maintenance company) has equipped the Pendolino Fleet with ROCM (improving reliability by up to 30%), the CP Porto 34 strong commuter fleet (energy savings up to 20%), and also the CP Freight fleet of elderly diesel locomotives which, with new on board sensors, now have increased performance and life expectancy;
» Norwegian State Railways (NSB), which has equipped nearly 300 trains with information distributed to more than 200 people from different departments;
» NS (the Dutch state rail operator), which has severe over crowding problems, has used a link to the output of passenger counting systems to direct people to less congested parts of the train.
Crossing the safety divide
Providing a connectivity service for commercial and monitoring purposes is one thing, but transmission of operational messages, for instance Movement Authorities to trains, is another ball game. It begs the question as to when an application becomes safety related or even safety critical.
Where does GSM-R sit in this debate?
It is certainly a fundamental part of ERTMS and, without reliable communication links, trains will be brought to a halt. In that sense, GSM-R is safety related. Only if radio messages are corrupted to give legitimate but invalid false information to ETCS can it be regarded as safety critical.
As mentioned earlier, any successor system will almost certainly not have dedicated bandwidth but will this scenario worsen the safety factor? It could be that availability is improved, thus providing a more reliable service. This can happen if multiple radio links are used where a message may be delivered by any one or more of these, the resulting redundancy giving a higher availability of a data ‘path’ to and from the train, something that cannot be achieved by reliance on a single technology.
The answer may come from North America, where Nomad has already provided the track-to-train control communications for the PTC (Positive Train Control) systems. PTC is a near-equivalent to ERTMS and it is now common practice for quasi-public networks to be used for PTC operational critical messaging. The safety of the message is assured by using authentication techniques plus methods for ensuring integrity and timestamps.
Multiple transmission modes and channels, including private radio networks, cellular mobile and Wi-Fi, are planned. However, these will have a closely managed QoS (Quality of Service) in a message server to support message prioritisation, delivery assurance and defence against denial of service.
Whilst the USA is predominantly a freight rail operation for long distances, passenger train usage in urban and city areas is on the increase, so speed and reliability are as important as in Europe. Public IP radio networks seem to be the way forward for operational messaging and Nomad Digital will no doubt be in the forefront of developing the associated systems to support this.
All in all, train connectivity has come a long way over the last decade, progressing from a ‘nice to have’ to becoming an essential part of a modern day train operation. Expectations must, however, be realistic and must align with radio frequency, bandwidth and coverage constraints.
Companies such as Nomad Digital, considered to be the pioneer and leader in the business, must be prepared to invest in network enhancement so as to ensure the planned service levels can be delivered reliably and consistently. What is clear is that train connectivity in all its facets is here to stay.
Why not use powerline transmission? OLE + Pantograph seems like a useful path to transmit data to the Train.
This is an interesting idea. The technology for this already exists and is even used by ISPs in Europe; the technology is known as Broadband over Power Lines (BPL). However, I see one major problem for the OLE-Pantograph interface. Despite it seeming that the two are attached at all times, the Pan often bounces on and off the contact wire – especially at high speed. This makes the connection unstable and causes a large amount of arcing – you can often hear a cracking sound when stood on a platform as an electric train passes by at speed. While the trains are built with huge capacitors to even the power supply to traction, the arcing causes a massive amount of interference that is likely to cause a lot of noise for any BPL. So, without further information on how this can be mitigated, if at all possible, then I don’t think this would work or be very effective.
Thank you for your thoughtful response. I believe that signal protocol could get around the interruptions/corruptions caused by breaks in the contact.
One of the factors determining the useful bandwidth would be the ratio of time in contact to the time lost to discontacts.
Radio transmissions are also subject to interference and interruption.
This is an interesting idea. The technology for this already exists and is even used by ISPs in Europe; the technology is known as Broadband over Power Lines (BPL). However, I see one major problem for the OLE-Pantograph interface. Despite it seeming that the two are attached at all times, the Pan often bounces on and off the contact wire – especially at high speed. This makes the connection unstable and causes a large amount of arcing – you can often hear a cracking sound when stood on a platform as an electric train passes by at speed. While the trains are built with huge capacitors to even the power supply to traction, the arcing causes a massive amount of interference that is likely to cause a lot of noise for any BPL. So, without further information on how this can be mitigated, if at all, then I don’t think this would work or be very effective.