AINews & Analysis

Steam Engines to Driverless Trains: What Next for India?

driverless train

With discussions around automation and artificial intelligence (AI) being the flavor of the day, we hear so much about driverless cars, trucks, and buses. But seldom do we talk about driverless trains that have been around for over 40 years even as there’s a growing buzz around this topic in India too–with good reason.

India has announced it will have its own first-ever driverless train running in the country, with Prime Minister Narendra Modi launching its operations on Delhi Metro’s 37 kilometre-long Magenta Line on 28 December.

To be sure, unlike driverless cars that are powered with AI technologies, India’s driverless train is a highly automated machine but not an AI driven one as some rail systems around the world. India’s new driverless trains have been equipped with Unattended Train Operations (UTO) and Communication Based Train Control (CBTC) signaling systems. They also come with energy-efficient subsystems like LED lighting, an enhanced braking mechanism, and other typical metro train features like air conditioning.

How do we define a driverless train?

Like the levels of driverless cars defined by SAE International, the International Association of Public Transport has notified five Grades of Automation (GoA) for these trains.

Level 0, for instance, relies entirely on a train driver to manage the system safety. Achieving the Level 1 would imply that a train driver controls starting and stopping, operation of doors and handling of emergencies or sudden diversions. However, s/he has the benefit of an Automatic Train Protection, or ATO, system.

five Grades of Automation (GoA)Source:

The next Level 2 is a semi-automatic train operation (STO) where stopping is automated but a driver in the cab starts the train, operates the doors, drives the train if needed and handles emergencies. A Level 3 driverless train operation (DTO) implies that starting and stopping are automated, but a train attendant operates the doors and drives the train in case of emergencies. As you may have already guessed, the top level, or Level 4, is a fully-automated train without any human staff.

According to UITP, there were 64 automated metro trains ran in 46 cities around the world in 2019. But automation is no stranger to Rail transportation. London Underground’s Victoria line, for instance, was the first ATO line in the world that was launched way back in 1967. However, it was Kobe New Transit’s Port Liner, which commenced operations in 1981, that was the world’s first automated mass-transit unmanned railway network.

Lille Metro in northern France has a similar network. Likewise, the Paris Metro Line 14, Line 4 of the Sao Paulo Metro, Barcelona Metro line 9, Copenhagen Metro and the Sydney Metro are examples of Level 4 automated trains. Incidentally, 22 six-coach driverless trains that were assembled in Andhra Pradesh by Alstom under the Make in India initiative, began running on the Sydney Metro line in Australia last May.

Other examples include the Govia Thameslink Railway that launched a mainline self-driving train service in London two years ago. A year later, in 2019, Japan tested a self-driving railway vehicle on East Japan Railway Company’s (JR East’s) Yamanote Line, which connects major stations in Tokyo such as the Tokyo Station, Shibuya Station, and Shinjuku Station. This May, French public transport company RATP said it was preparing one of the oldest routes of Paris Metro for ATOs.

And in August 2019, three locomotives and 30 loaded wagons carrying 4,725 trailing tonnes were moved without human intervention along a 48-mile test track in Colorado.

Do costs outweigh benefits?

ATOs do have economic advantages that include “…improved operational safety and lower rail fatalities with staff savings as large as 70%, and more than 30% improvement in energy efficiency. The space previously occupied by train operators could be utilized to accommodate more passengers”.

But would the operational costs outweigh the huge initial investments in these automated rail systems? While transportation experts and governments would like us to believe that the answer is a firm ‘Yes’, unions expectedly would beg to differ. According to a 26 October 2020, article, a presentation said to have been prepared for senior Transport for London (TfL) managers claimed that introducing driverless trains across London Underground would cost £7 billion and would represent “poor value for money”.

That said, driverless trains won’t replace drivers in a country like India or for that matter anywhere in the world in a hurry. The reason: there are quite a few hurdles that need to be overcome before we can even think about taking driverless trains mainstream.

To begin with, ATOs mandate closed systems. The lines need to be elevated and tracks must be isolated from ordinary roads so that there are no trespassers–be they animals or humans. People and vehicles almost never enter the tracks, and there are no railroad crossings in a bid to avoid accidents. The Chuo Shinkansen of Central Japan Railway Company (JR Central), slated to open in 2027, is a good example of such an ecosystem. Further, the Chuo Shinkansen runs at a speed of 500 km/h on a Magnetic Levitation (Maglev) type linear motor railway.

This explains why India’s driverless train will be run on the metro line–at least in the interim period. The same, of course, will apply to hyperloops.

Heralding the future: AI-powered smart trains

AI technologies are already powering rail systems around the world. The Dubai International Airport, for instance, operates a fully autonomous train to transfer passengers from one terminal to another. It uses SelTrac, an automatic railway signaling technology from Toronto-based company Thales, to control the train autonomously. The Singapore Mass Rapid Transit Lines, currently the world’s longest automated metro system, is another case in point. This urban transit system has achieved Level 4 automation.

Wasay Rashid, ARC Advisory Group analyst points out in a blog that in the European Union, the first key step towards the introduction of ATO and AI solutions in rail transport is the deployment of the European Rail Traffic Management System (ERTMS) that already provides trains with a driver assistance system.

“Today, we’re seeing AI being used in rail applications to improve train scheduling, manage train speeds, avoid accidents, predict delays, enhance asset management, and more. These AI applications help ensure public safety, deliver customer value, and optimize overall rail management and operations. In this manner, the technology is helping reverse the trend for rail transport to lose market share to other modes of transportation,” he concludes.

It’s another matter, altogether, that around three years back, India’s Union minister for road transport and highways, Nitin Gadkari, asserted that driverless cars will not be allowed in India because the government is not going to promote any technology that comes at the cost of jobs.

In the case of driverless trains, the government has no objection to the technology. However, India’s railway worker unions may have a different point of view.

(Leslie D’Monte is Publishing Director, Trivone)

Leave a Response