A report in the Australian Financial Review and Ten News yesterday suggests that Transport for NSW wants the Northwest Rail Link (NWRL) built with the potential for driverless trains to run on the line. Such a system exists in many parts of the world, such as Vancouver, London or Dubai.
The main advantage of this is the low cost, as you no longer have to pay a driver to operate the train. This benefit would be seen most clearly during the off-peak, where you can maintain frequent services thanks to the lower variable costs (most costs are up front and fixed: track, stations and trains). Jarrett Walker describes how on the Vancouver driverless metro: the SkyTrain (portions of which are driverless) you never have to wait more than 8 minutes for a train, no matter what time of the day or week it is, helping to maintain high patronage all day rather than just during peak hour. The reason it can do this is the efficiency that comes from the low operational cost.
The news piece from Ten News presents 2 arguments against driverless trains: cost and safety.
The first is by Shadow Transport Minister Penny Sharpe, who says that this is just a cost cutting exercise. That is correct and is simple economics – the lower the variable cost of running one train, the more trains you can run for the same amount of funding. You could argue that the government should just increase funding to achieve this, but then you could have increased services even more. Either way, lower cost is a good thing, not a bad thing.
The second is by Bob Nanva of the Rail, Tram and Bus Union, who is concerned about the safety of not having a driver on the train. He cites the case of a driverless train on the London Underground that went 6km before it was stopped, almost hitting another passenger train in the process. There are 2 problems with this assertion. First, human error has been responsible for 14 deaths in Sydney in the last 15 years due to the train accidents at Waterfall and Glenbrook, neither of which may have happened with an automated system that takes human error out of the equation. Second, the driverless train was actually an engineering carriage without a driver that was being towed by another train when the coupling between them broke. Despite the towing train having a driver,which under this argument should have detected the incident and dealt with it, this was not noticed until signallers detected it and resolved the potential disaster. In other words, it was the driverless controls in place that prevented any escalation of the situation, not the presence of a driver. I would therefore argue that a driverless system is actually more safe, not less safe, than one with drivers.
Ultimately, it’s important to remember that Premier Barry O’Farrell’s response to questions on this led to him responding that “it’s not a prospect that’s been put to me” and that Transport Minister Gladys Berejiklian stated that they “are planning for the trains on this important rail link to have drivers”, suggesting that this is just Transport for NSW making sure that the NWRL is future-proofed for driverless trains should a NSW Government ever choose to go down that path in the future. But it leaves the option open to that, rather than being the start of a process to make it happen.
Realistically, the government would first have to remove guards from trains, arguing that one person should now be capable of performing the duties of driver and guard, before making the move to trains that are no longer staffed on a permanent basis. And this itself would be a huge fight with the union. When and if it does that, then we can start speculating on whether driverless trains might become a reality.