Should trains go driverless?

Fully driverless underground and overground trains are becoming the norm in many cities around the world, including several in Europe.

Interestingly enough, the world seems to be reluctant to fully go driverless on mainline rail networks. However, with the development of the Docklands Light Railway, many argue it is the right time to start looking in fully automating the operation of the entire UK rail network.

So, what’s the right way forward?

Automation is the natural evolution of rail systems, argue Hugh & Max

The future of road travel is most certainly autonomous cars, so why not apply this concept to an environment more controlled and predictable than roads, that is mainline railway networks? Over the years, we have seen driverless trains successfully introduced to metro systems worldwide, and it is now time to take this technology and apply it to mainline networks too.

Some may argue the safety of fully automated trains since drivers have accumulated years of expertise and replacing them could have unpredictable consequences. However, it turns out these vehicles might actually be safer than conventional trains, as they remove one of the main causes of rail accidents – human error.

One of the country’s worst rail accidents in recent times was the Ladbroke Grove crash in 1999, in which 31 were killed and more than 520 injured. The cause of this incident was a so-called “signal passed at danger” (SPAD), that is the train passed a stop signal when it wasn’t authorised to do so. SPADs are common causes of accidents, as the signalling system works on the principle of the driver establishing visual contact with the trackside or overhead signal. The driver may miss the signal due to distraction, fatigue, or misunderstanding. Driverless trains completely eliminate this aspect, as there isn’t any need for a visual signalling system, as the train constantly evaluates the safety to proceed through contact with nearby trains and trackside sensors.

Not all accidents are a result of human error though – another major cause of rail accidents is objects on the track.

Mainline railway networks by their nature are open systems, meaning railway lines are not isolated from public space. Collisions between trains, people and objects are therefore almost inevitable; and with modern high speed trains, there is a significant risk of death associated with trespassing on the track.

For example, if a person finds themselves on a track with a conventionally-controlled train travelling towards them at high speed, there would hardly be any reaction time for the driver to slow the train. The situation would be very similar with a driverless train, even if the train had a system to detect a person on the line ahead.

Furthermore, fully automated trains will increase the capacity of railways, thus addressing the issue of overcrowding and encouraging more passengers to commute by train rather than car, hence reducing their carbon footprint. 

Despite trains being safer than other modes of transport, deaths, despite their almost negligible number, remain a consequence regardless of whether trains are automated or not. Going driverless though, would eliminate the possibility for human error, the leading cause of accidents. The improvements in safety, service and reliability turn driverless trains into the lynchpin of the rail renaissance.

Automation threatens the core aspect of rail infrastructure: reliability, argue Sotiris & Tochi

There is an important reason why pilots, captains and drivers are still around, while we boast the technical capacity to fly drones and command a rover on Mars remotely. And that reason is the distinct human need to feel safe. Having human beings in charge, makes commuters feel secure and that they are being taken care of. This, although merely a feeling, enables people to trust public transport more and therefore make frequent use of it.

Replacing drivers presents thus the risk of sacrificing the progress we’ve made over the years to encourage people to leave their cars at home. That, combined with increased susceptibility of automated systems to cyber attacks, means automation could make public transport quite unpopular. And we really cannot afford to let that happen.

Eliminating the human factor will also be a blow to the British long-standing tradition of making and operating rolling stock. Rail, having been invented in England, helped spread the wonders of the Industrial Revolution to the rest of the world and remains a source of great pride for Britons. If you think that’s an exaggeration, then you should take a look at the huge crowds around the nation’s tracks last week waving at the Flying Scotsman.

Another key aspect is the unfathomable cost of mainline automation. Underground trains operate in closed networks, equipped with endless sensors, CCTV and warning signs to block people (or things) from the tracks, thus making automation easy to implement. The largely rural UK mainline network, on the other hand, is too open to external disruptions, such as physical disasters or people finding themselves on the rails intentionally or unintentionally. This inevitably means that we need a huge amount of expensive sensors to enable the train to make the right decision in an emergency. And that’s perhaps the toughest, costliest part.

Let’s assume there is a person on the rails within braking distance. Will the train decide to brake and save the person, risking collision with another train travelling in high speed just behind it? Or will it choose to kill said person, in the hope not to disrupt the punctual service and/or endanger its passengers? How will the train be able to assess the level of the emergency presented? Will the sensors be able to tell the difference between a fox, a log and a person on the rails and act differently?

Nobody has an idea of what the answers to these questions may actually be.

Long story short, driverless trains will cause a high level of mistrust to the public due to driver lay-offs, lack of human presence and the potential for a remote, electronic security breach. They will also bear an inexorably high cost that won’t in any way guarantee increased safety or better service.

60: Hugh Garrett-Allen, Max Roberts, Sotiris Kopatsaris, Tochi Emenike


8 thoughts on “Should trains go driverless?

  1. Automation should be encouraged but not for trains to be driverless. Though with sophisticated voting system but a spurious trip on a driverless train could lead to disaster -high fatality.


  2. I think that if there are no major issues related to the operation of trains by drivers at the moment, then there is no need for automation. Automation might even create more problems than it could solve, such as driver lay-offs, at a time when global unemployment is projected to grow. Instead, investing resources towards improving the existing systems in terms of safety and efficiency (eg preventing SPAD incidents) could be more beneficial.


  3. Automated trains definitely has its pros as it reduces human errors which is one of the biggest cause of rail accidents but from an economic point of view, job loss as well as the high cost of implementing automation services presents a negative impact.
    One way forward is to have a quasi automated system. This way, drivers can keep their jobs, the cost of implementing automation is reduced as manual operation will be used in various areas and automated operation is used in critical areas.


  4. My major objection to fully driverless automated trains is the perception that handling complete power to a machine poses a threat to the safety of its users . There’s this feeling That gives a bit of comfort when you know that someone with experience who can react to unusual and unexpected situations is in charge. machines are basically programmed in one way. So if things don’t go the way the machine has been programmed to, it shuts down and this can be at the detriment of the users. Unlike when an actual person with experience is in charge .


  5. It’s great that technology is evolving but I would rather not go on an automated train. An automated train is programmed. I’d personally feel safe knowing there’s someone to handle things if something goes wrong. There’s little or no guarantee that an automatic train would do same.


  6. Full automation of trains sounds like a brilliant idea with lots of benefit such as the removal of human errors, reduced overcrowding and hence reduced carbon footprint by cars but author believes that these do not compensate for the disadvantages this idea poses.
    For one, fully automating trains would mean a large retrenchment of train drivers and some other personnel leading to an increase in unemployment rate which could invariably lead to increased depression and crime. Secondly, drivers have expertise from years of experience and could deal easily with unforeseen circumstances for instance, a fully automated train could easily be hijacked by terrorist leading to grave catastrophe, this would not be the case if the train was being controlled by a driver who would be able to contact the necessary authorities and do some meanderings to abort the hijacking. This expertise and humanity of the driver would also come into play if an individual was caught on the rail road, he/she would be in a better position to know how to save this individual if possible, an automated train on the other hand, has no feelings or ethical considerations and would plow on since getting to its destination on time is its forefront objective. Thirdly, with the present state of the economy, full automation of trains is not cost effective since expensive infrastructure would have to be put in place especially in the rural areas to make this automation possible, fourthly, full automation of trains would make people wary since humans like the feeling of knowing another human with experience and expertise is in charge of their safety and not just a set of computerized equipment; furthermore, this full automation would take away a source of British pride (since rails invented I Britain were the source of the industrial revolution).
    In conclusion, the author believes that the idea of full automation of trains whilst a great and beneficial idea is one that is still quite premature; alternatively, just like airplanes, trains could be made more automated but still need the presence of an expert human.


  7. I agree that trains should be automated. This is because humans can easily make mistakes; alcohol, tiredness, injury, drugs, mental instability etc. Any one mistake can destroy lives. Another benefit is the cost saving – the initial capital will be costly but the savings will be realised shortly. Services should be more reliable and efficient – errors should be easier to detect and if they are rectified before they become an issue, this should provide an overall better service. When a service provided does what it sets out to do, people do not hesitate to spend their money. No chance of strikes and so productivity in time and money is greatly increased


  8. I understand the arguments for automating the rail network, but I don’t support it. Aside from the cost of implementation and vulnerability of the system to hackers, there’s going to be a quite a spike in unemployment rate because the drivers will be out of jobs. And that can’t possibly be good for the economy, especially when so much money would have to go towards implementing it in the first place.


Leave a Reply

Fill in your details below or click an icon to log in: Logo

You are commenting using your account. Log Out /  Change )

Google+ photo

You are commenting using your Google+ account. Log Out /  Change )

Twitter picture

You are commenting using your Twitter account. Log Out /  Change )

Facebook photo

You are commenting using your Facebook account. Log Out /  Change )


Connecting to %s