Uber is an on-demand taxi service you can book and pay for on your mobile phone. Writing for the Guardian last year, our own Mark Wallace was impressed with how it all worked. Traditional cabbies were rather less impressed and held protests in cities across Europe.

Clearly, the taxi drivers don’t like the idea of a multinational company muscling-in on their turf. However, what they really should be worried about isn’t some extra competition, but being replaced by robots. According to John Biggs in TechCrunch, Uber is “building a robotics research lab in Pittsburgh”:

“Sources tell us Uber is hiring more than fifty senior scientists from Carnegie Mellon as well as from the National Robotics Engineering Center, a CMU-affiliated research entity. Carnegie Mellon, home of the Mars Rover and other high-profile robotics projects, declined to comment at this time, as did scientists mentioned by our source. Uber has ‘cleaned out’ the Robotics Institute, said the source…

“In the past, Uber CEO Travis Kalanick has said he would replace human drivers with self-driving cars.”

Driverless vehicles are now being taken very seriously by business and government alike. Only last week, the UK Government launched its ‘Pathway to Driverless Cars’.

Commenting on the document for Prospect, Emran Mian gives it his qualified approval:

“…on the whole the plan is informed by a sophisticated understanding of the car industry, the challenges facing ‘autonomous vehicles,’ and what government can do about them.”

In particular, he notes the “level headed” recognition that the “technology will develop gradually, perhaps taking as much as 20 years before it starts to transform anything.”

Twenty years might seem like a long time in politics; but as Mian notes, this is the timeframe over which some very large infrastructure investments are being committed today. If new technology does start transforming transport over this period, shouldn’t we rethink some of our assumptions?

HS2 is an obvious example:

“Consider the new North-South rail link to deliver a high speed rail line between London and other major cities (at high cost.) What if by the time the line is built—roughly 20 years—driverless cars can make those journeys almost as quickly with much greater flexibility on departure and arrival times than rail can offer?”

Compared to ordinary cars, driverless vehicles come much closer to rivalling the passenger experience that trains provide:

“One of the other advantages of rail at the moment is that people can work on trains whereas they can’t in cars. That advantage disappears if the driver doesn’t have to, well, drive—they can be reading, emailing or videoconferencing at the same time, potentially while enjoying greater privacy and comfort than a train can offer.”

Of course, key differences remain between the two modes of transport. Whether automated or not, road vehicles are slower than trains and don’t have the same freight capacity. No matter how advanced driverless technology becomes, rail will fulfil a role that road transport can’t.

However, that leaves the question as to whether the money required for HS2 couldn’t be used more productively elsewhere on the rail network. The ConservativeHome manifesto makes the case for scrapping HS2 and putting the money into improved east-west rail links across the cities of the North.

Intelligently controlled, driverless vehicles would only increase the potential of such a network – by seamlessly extending its reach from the major cities out to every town and neighbourhood across the region.

For too long, England’s ‘northern powerhouse’ – as George Osborne calls it – has been held back by its outdated infrastructure. With new technology multiplying the possibilities for a truly integrated transport system, is the North really best-served by an expensive, but otherwise unremarkable, train line to London?